# Complete AI Business: full text > The public pages of https://completeaibusiness.com as Markdown, in one file. The short index is https://completeaibusiness.com/llms.txt. Customer plans are private and not included. --- --- title: "Complete AI Business | Your business AI-ready in 7 days" description: "We make businesses with up to 21 people AI-ready in 7 days for $7,700, fully online over Zoom: the right AI platform, every employee's AI set up, training for the whole team and your main automations built." url: https://completeaibusiness.com/ --- For teams up to 21 people · 100% online, over Zoom # Your business AI-ready in 7 days. One fixed price: $7,700. We pick the right AI platform for your business, set up every employee's AI, train your whole team and build the automations that take the repetitive work off your plate. Seven online sessions in seven days, all over Zoom, and we guide you through every step. [Get your free AI plan](https://completeaibusiness.com/questionnaire) [See the case studies](https://completeaibusiness.com/case-studies) - ✓ Free AI plan in about 5 minutes - ✓ All sessions online, over Zoom - ✓ ChatGPT, Claude, Copilot or Gemini - ✓ Money-back guarantee on the initial call **Your 7 days** One Zoom session a day Day 1 · Whole team on Zoom, 60 min ### Kick-off and discovery We interview you and your team, map the work and the tools you use, and agree what AI-ready means for your business. - Short interviews per role - Tools and data check - Your top time drains, ranked AI-ready > I was sceptical, but within a week my team stopped dreading the monthly bank reconciliation. We now get it done in under two hours instead of two days. Complete AI Business really understood our specific accounting workflows. **Jan van der Berg**, Managing Partner, Van der Berg Accountants (14 people) [Read how they did it →](https://completeaibusiness.com/case-studies/accountancy) We use fictive names for privacy reasons. These are illustrative examples. ![Training day on Zoom: the client's team is at their own workplace on a Zoom call, with two accountants building a reconciliation automation in Copilot on a large screen while colleagues take notes on laptops, and the consultant visible on a video screen.](https://completeaibusiness.com/img/cases/accountancy-2.webp) ### 100% online, over Zoom Every session is a Zoom meeting. Nobody travels: your team joins from their own desk, or together from your meeting room. ### Your screen, your tools You share your screen and we set things up together in the tools you already use, so it works on your real work from day one. ### Wherever your team is One office, several locations or people working from home: it makes no difference, because it all happens online. What you get ## Everything your team needs to work with AI. Set up with you, not handed over. You don't need to know which tool does what. We advise, set up, train and build, and your team learns by doing it on their own work. ### Advice on the right AI platform ChatGPT, Claude, Copilot or Gemini: we advise the one that fits your tools, your team and your data rules. Plus a clear list of what to automate first, and what to keep human. ### An AI training plan for every person Online, per person and per job, so the accountant learns other things than the planner. Built on the courses and certificates of Complete AI Training. From 700+ courses with certificates ### Every employee's AI, set up and ready On a one-on-one Zoom call with each person, screen shared, we set up their AI for their job: projects, instructions and connected apps. With written guides and short videos, so it sticks afterwards. ### Your main automations, built The repetitive work that eats your week goes to AI agents that run every day: inboxes, quotes, reports, planning, follow-ups. Connected to the tools you already use. ### 7 online consultancy sessions One Zoom session a day for seven days, from the kick-off to the handover. Short, focused and on your real work, so your business keeps running while it happens. ### Live training and custom videos One live training on Zoom for the whole team, and short training videos made for your business that show each person what their AI can do in their job. For teams up to 21 people. Bigger team? We add a fee per extra employee and agree it with you before we start. How the 7 days work ## One session a day. AI-ready on day seven. Each day has one Zoom session with one clear goal. Most people on your team join two or three: the kick-off, their own setup and the live training. Everyone joins from their own desk. 1. Day **1** ### Kick-off and discovery Whole team on Zoom, 60 min We interview you and your team, map the work and the tools you use, and agree what AI-ready means for your business. 2. Day **2** ### Platform advice Owner and key people, on Zoom We advise the AI platform that fits your business and set up the accounts, data settings and shared company knowledge. 3. Day **3** ### Everyone's AI, set up One-on-one Zoom call, every employee On a one-on-one Zoom call with each person, screen shared, we set up their AI for their job: projects, instructions and connected apps, with a short written guide. 4. Day **4** ### Live team training Whole team, live on Zoom A live Zoom training for the whole team. Then every person starts an online training plan built for their job, with certificates. 5. Day **5** ### Build the main automations Consultant and process owners, on Zoom We build the automations that take the most time off your week, connected to the tools you already use. 6. Day **6** ### Test, fix and film The people who use it, on Zoom Your team uses everything on real work. We fix what is rough and record short training videos made for your business. 7. Day **7** ### Handover: AI-ready Owner, then the team, on Zoom Documentation, videos and a 90-day plan. We go through the AI-ready checklist together, and your team takes it from there. Pricing ## One fixed price. Start small if you like. Start with the initial call and planning, or go straight to your 7 days. After that, a retainer keeps it running, if you want one. Start here ### Initial call and planning **$770** one-off We go through your free AI plan together, sharpen it with you, and hand you a written implementation plan for your business. - Strategy call with a consultant, on Zoom - Written implementation plan - Platform and automation advice - Clear scope for your 7 days ✓ **Worth more than you paid, or your money back.** If the call and plan are not worth more than $770 to you, we refund you. [Book the initial call](https://completeaibusiness.com/questionnaire?start=call) Done with you ### AI-ready in 7 days **$7,700** one-off, up to 21 people* Everything to make your business AI-ready, in seven working days. - 7 online consultancy sessions, over Zoom - Advice on the right AI platform and what to automate - Every employee's AI set up, with documentation - An online AI training plan for every team member - A live Zoom training and custom training videos - Your main automations, built and tested *Bigger team? We add a fee per extra employee, agreed before we start. [Start my 7 days](https://completeaibusiness.com/questionnaire?start=full) Optional ### Monthly retainer from **$770** a month, or every other month We keep your AI running and improving after the 7 days, so you never fall behind. - Maintenance and managing your AI infrastructure - Personal training for your team, on Zoom - Consultancy updates when AI changes - New automations as your business grows [Ask about the retainer](https://completeaibusiness.com/questionnaire?start=retainer) Case studies · 9 industries ## What AI-ready looks like in a business like yours. Nine example engagements: the team, the advice, the 7 days, the automations and the hours they got back. Professional Services and Consulting ### Van der Berg Accountants Precision bookkeeping and tax advice for Dutch SMEs **45h** back per week **4 wk** to pay back **14** people [Van der Berg Accountants](https://completeaibusiness.com/case-studies/accountancy) Marketing, Advertising and Creative Agencies ### Juniper & Hay Brand identity and content for ambitious consumer and B2B brands **38h** back per week **5 wk** to pay back **12** people [Juniper & Hay](https://completeaibusiness.com/case-studies/creative-agency) Healthcare Providers and Clinics ### Sunset Ridge Dental Generations of healthy smiles for Austin families since 1995. **49h** back per week **4 wk** to pay back **18** people [Sunset Ridge Dental](https://completeaibusiness.com/case-studies/dental-clinic) Legal Services ### Vandermeer & Peeters Advocaten Employment and contract law for small and mid-sized businesses **30h** back per week **4 wk** to pay back **9** people [Vandermeer & Peeters Advocaten](https://completeaibusiness.com/case-studies/law-firm) Real Estate and Property Management ### LarViva Imobiliária Helping Lisbon families buy, sell and rent homes since 2008. **36h** back per week **6 wk** to pay back **11** people [LarViva Imobiliária](https://completeaibusiness.com/case-studies/real-estate) E-commerce and D2C ### EcoCulinar Sustainable kitchenware for modern European homes. **25h** back per week **9 wk** to pay back **8** people [EcoCulinar](https://completeaibusiness.com/case-studies/online-store) Construction and Engineering ### De Vries Renovatie & Installatie Rotterdam renovation and installation contractor for homes and businesses **45h** back per week **4 wk** to pay back **21** people [De Vries Renovatie & Installatie](https://completeaibusiness.com/case-studies/renovation) Hospitality, Hotels and Restaurants ### The Thistle & Thorn 34-room boutique hotel with a seasonal Scottish restaurant in Edinburgh's Old Town **37h** back per week **7 wk** to pay back **19** people [The Thistle & Thorn](https://completeaibusiness.com/case-studies/boutique-hotel) Insurance ### Bayview Insurance Brokers Independent insurance advice for Australian small businesses. **31h** back per week **6 wk** to pay back **10** people [Bayview Insurance Brokers](https://completeaibusiness.com/case-studies/insurance-broker) We use fictive names for privacy reasons: the companies, people and quotes are fictive, and each case shows how a typical 7-day engagement runs in that industry. [All case studies →](https://completeaibusiness.com/case-studies) How to start ## Three steps. The first two are free. 1 About 5 minutes ### Answer the questionnaire Your business, your team and the roles in it, the tools you use, and what you like and don't like doing. 2 A few minutes later ### Read your free AI plan We research your business online and write a first plan: the platform, training per role, everyone's AI setup and automation ideas. 3 When you're ready ### Start your engagement Book the initial Zoom call and planning for $770, money back if it's not worth it. Or start your 7 days right away for $7,700. The Complete AI family ## Learn it, hire it, or get your whole business ready. Same team, same way of working. Complete AI Business works with its sister brands: your team's training plan uses Complete AI Training, and single agents can run on Complete AI Agents after your 7 days. Learn AI for your job AI kits, 700+ courses and certificates per job. Every person on your team gets a training plan from here. [Open](https://completeaitraining.com/) Hire one AI agent One agent for one boring process, set up and run by a human consultant. Good for adding automations later. [Open](https://completeaiagents.com/) ![Complete AI Business](https://completeaibusiness.com/assets/family/business.svg) You are here · Your whole business The platform, everyone's AI, the training and your main automations, in 7 days for $7,700. Questions ## Frequently asked **Is everything online?** Yes. Every session happens in a Zoom meeting: the kick-off, the one-on-one setup with each employee, the live training and the handover. You share your screen and we work in your own tools together. Nobody travels, and it works wherever your team is. **Which AI platform will we use?** The one that fits your business, usually ChatGPT, Claude, Microsoft Copilot or Google Gemini. We look at the tools you already use, how your team works and what your data rules are, and advise one. Your free AI plan already includes a first recommendation. **Does my team need to be technical?** No. We guide every person through every step, join them on a Zoom call for their own setup, and leave written guides and short videos made for your business. If you can join a Zoom meeting, you can do this. **What if we have more than 21 employees?** We still do it. We add a fee per extra employee and agree it with you before we start, so there are no surprises. **How does the money-back guarantee work?** It applies to the initial call and planning ($770). If you don't think the call and the written plan are worth more than you paid, tell us and we refund you. **What happens after the 7 days?** Your team runs it, with the documentation, videos and 90-day plan we leave behind. If you want us to keep it running and improving, add a retainer from $770 a month, or every other month. **How much time does it take from my team?** There is one Zoom session a day. Most people join two or three: the kick-off, their own setup and the live training. The people who own a process also join the day we build its automation. **What about our data?** We set up business accounts with the data settings your platform offers for companies, agree with you what goes into the AI and what stays out, and write it down in your documentation. ## See what AI can do for your business. In about 5 minutes. Answer the questionnaire and get your free AI plan: the platform, the training, everyone's setup and the first automations. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "Case studies | Complete AI Business" description: "Nine example engagements in nine industries: the team, the advice, the 7 days, the automations and the hours they got back." url: https://completeaibusiness.com/case-studies --- Case studies · 9 industries # AI-ready in 7 days, in nine different businesses. An accountancy, a dental practice, a renovation contractor and six more. For each one: the team, what they told us, our advice, their 7 days, the automations we built and what it saved them. **We use fictive names for privacy reasons:** we don't publish who our clients are. The companies, people and quotes in these case studies are fictive; each one shows how a typical 7-day engagement runs in that industry. ### Hours back per week, per business From daily AI use and the automations built in the 7 days. [Sunset Ridge Dental 18 people 49h](https://completeaibusiness.com/case-studies/dental-clinic) [Van der Berg Accountants 14 people 45h](https://completeaibusiness.com/case-studies/accountancy) [De Vries Renovatie & Installatie 21 people 45h](https://completeaibusiness.com/case-studies/renovation) [Juniper & Hay 12 people 38h](https://completeaibusiness.com/case-studies/creative-agency) [The Thistle & Thorn 19 people 37h](https://completeaibusiness.com/case-studies/boutique-hotel) [LarViva Imobiliária 11 people 36h](https://completeaibusiness.com/case-studies/real-estate) [Bayview Insurance Brokers 10 people 31h](https://completeaibusiness.com/case-studies/insurance-broker) [Vandermeer & Peeters Advocaten 9 people 30h](https://completeaibusiness.com/case-studies/law-firm) [EcoCulinar 8 people 25h](https://completeaibusiness.com/case-studies/online-store) Professional Services and Consulting ### Van der Berg Accountants Precision bookkeeping and tax advice for Dutch SMEs. Utrecht, the Netherlands. “I was sceptical, but within a week my team stopped dreading the monthly bank reconciliation.” **45h** back per week **$113,900** value per year **4 wk** to pay back [Van der Berg Accountants](https://completeaibusiness.com/case-studies/accountancy) Marketing, Advertising and Creative Agencies ### Juniper & Hay Brand identity and content for ambitious consumer and B2B brands. Manchester, United Kingdom. “We used to lose Fridays to report formatting and cutdowns.” **38h** back per week **$87,400** value per year **5 wk** to pay back [Juniper & Hay](https://completeaibusiness.com/case-studies/creative-agency) Healthcare Providers and Clinics ### Sunset Ridge Dental Generations of healthy smiles for Austin families since 1995. Austin, Texas, USA. “I used to spend three hours every Friday building that production report.” **49h** back per week **$94,700** value per year **4 wk** to pay back [Sunset Ridge Dental](https://completeaibusiness.com/case-studies/dental-clinic) Legal Services ### Vandermeer & Peeters Advocaten Employment and contract law for small and mid-sized businesses. Antwerp, Belgium. “The first draft used to take me most of a morning.” **30h** back per week **$103,500** value per year **4 wk** to pay back [Vandermeer & Peeters Advocaten](https://completeaibusiness.com/case-studies/law-firm) Real Estate and Property Management ### LarViva Imobiliária Helping Lisbon families buy, sell and rent homes since 2008. Lisbon, Portugal. “I was worried my team would sound like robots, but the system we built actually sounds more personal because it remembers details I forget.” **36h** back per week **$66,200** value per year **6 wk** to pay back [LarViva Imobiliária](https://completeaibusiness.com/case-studies/real-estate) E-commerce and D2C ### EcoCulinar Sustainable kitchenware for modern European homes. Copenhagen, Denmark. “Before, our support team was drowning in 'where is my order?' emails.” **25h** back per week **$43,700** value per year **9 wk** to pay back [EcoCulinar](https://completeaibusiness.com/case-studies/online-store) Construction and Engineering ### De Vries Renovatie & Installatie Rotterdam renovation and installation contractor for homes and businesses. Rotterdam, the Netherlands. “The AI took the paperwork off my desk.” **45h** back per week **$93,200** value per year **4 wk** to pay back [De Vries Renovatie & Installatie](https://completeaibusiness.com/case-studies/renovation) Hospitality, Hotels and Restaurants ### The Thistle & Thorn 34-room boutique hotel with a seasonal Scottish restaurant in Edinburgh's Old Town. Edinburgh, United Kingdom. “Within a week, we turned our biggest time drains into automated processes that feel natural.” **37h** back per week **$54,500** value per year **7 wk** to pay back [The Thistle & Thorn](https://completeaibusiness.com/case-studies/boutique-hotel) Insurance ### Bayview Insurance Brokers Independent insurance advice for Australian small businesses. Melbourne, Australia. “I was sceptical about AI understanding insurance, but the proposal generator alone saves my brokers hours each week.” **31h** back per week **$68,400** value per year **6 wk** to pay back [Bayview Insurance Brokers](https://completeaibusiness.com/case-studies/insurance-broker) ## What would your 7 days look like? Answer the questionnaire and get a free first plan for your business, with the platform, the training and the first automations. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "Van der Berg Accountants: 45 hours a week back with AI" description: "How a 14-person accountancy and bookkeeping firm in Utrecht, the Netherlands got 45 hours a week back in 7 days with Microsoft 365 Copilot: the team, the…" url: https://completeaibusiness.com/case-studies/accountancy --- ![A normal Wednesday morning: accountants at their desks with dual monitors showing Excel and Exact Online, while the office manager sorts a fresh stack of client folders.](https://completeaibusiness.com/img/cases/accountancy-1.webp) Professional Services and Consulting · Utrecht, the Netherlands # Van der Berg Accountants got 45 hours a week back in 7 days A 14-person accountancy and bookkeeping firm in Utrecht serving small and medium businesses across construction, hospitality, and professional services. The team handles year-end accounts, payroll, and day-to-day bookkeeping, with a reputation for personal, jargon-free advice. A normal Wednesday morning: accountants at their desks with dual monitors showing Excel and Exact Online, while the office manager sorts a fresh stack of client folders. **14** people on the team **45h** back every week **$113,900** value per year at $55/h **4 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** Van der Berg Accountants and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Precision bookkeeping and tax advice for Dutch SMEs - **Customers:** SMEs in construction, hospitality, and professional services, typically 5-50 employees. - **Team:** 14 people - **Location:** Utrecht, the Netherlands Tools before we started Exact Online Microsoft 365 (Excel, Outlook, Teams, Word) Mailchimp TimeChimp Adobe Acrobat WhatsApp (client messaging) ### What they sell, share of revenue Bookkeeping and monthly reports 40% Annual accounts and tax returns 35% Payroll processing 15% Business advisory 10% The team ## 14 people, and what their AI now does for them Managing Partner Taught Copilot to draft client emails, summarise meetings, and pull key metrics from Excel reports for client calls. 1 Senior Accountant Set up Copilot in Excel to generate complex analyses from trial balance data and write first drafts of advisory notes in Word. 2 Accountant Configured Copilot to reconcile bank statement CSV files with Exact Online exports and flag missing transactions. 4 Bookkeeper Set up Copilot to extract vendor, amount and date from scanned receipts and enter them into a shared Excel workbook. 4 Payroll Specialist Taught Copilot to cross-check payroll calculations against statutory rates and draft the monthly payslip summary email. 1 Office Manager Set up Copilot to categorise client emails, schedule team meetings from chat threads, and generate onboarding checklists. 1 Administrative Assistant Set up Copilot to assemble engagement letters and privacy forms by filling in client details from a brief prompt. 1 The questionnaire ## What Jan van der Berg told us before we started Jan van der Berg, Managing Partner, filled in the same questionnaire you can fill in today. Biggest time drain Manually categorising bank transactions and chasing client documents. The goal Reduce time spent on low-value tasks so we can serve more clients without hiring, and improve work-life balance. Worries about AI AI might make errors in tax filings · Losing the personal touch with clients · Team resistance to new technology ### Keep human - Personal client advisory conversations - Finding tax optimisation opportunities - Mentoring team members ### Take it off our plate - Data entry - Reconciling discrepancies in spreadsheets - Following up on missing receipts - Filling in repetitive forms for tax authorities Our advice ## Microsoft 365 Copilot, and the three things to automate first Your team already works in Outlook, Teams, Excel and Word every day. Copilot integrates directly with these tools and can connect to your Exact Online data via Excel, so you automate without switching environments. Your tools today Exact Online Microsoft 365 (Excel, Outlook, Teams, Word) Mailchimp TimeChimp Adobe Acrobat WhatsApp (client messaging) We advise **Microsoft 365 Copilot** Fit 10/10 Set up for 14 people, connected to your tools AI takes over - Bank transaction reconciliation - Client onboarding document pack - Scanned receipt data extraction - Weekly team report compilation Stays human, AI only supports Tax optimisation advice Complex advisory conversations with clients Final review of all filings ### How each platform fit this business Microsoft 365 Copilot Chosen 10/10 Google Gemini for Workspace 5/10 ChatGPT Business 4/10 Claude Team 3/10 ### What we advised to automate first - Reconciling bank transactions with bookkeeping entries - Drafting standard client communications and meeting summaries - Extracting data from scanned receipts into Excel Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### AI readiness audit and team kick-off Entire team We met on Zoom to map current workflows, identify the highest-effort manual tasks, and agree on priority automation targets. The whole team joined to voice their pains and ideas. 2. Day **2** ### Individual AI setup sessions All 14 team members, one-on-one Each team member spent 30 minutes on a one-on-one Zoom call with screen sharing, configuring their Copilot, crafting custom instructions for their most repetitive work, and testing first prompts. 3. Day **3** ### Reconciliation and receipt automations 4 accountants, 4 bookkeepers On Zoom, accountants and bookkeepers built a bank statement matching automation in Excel and a receipt data extractor from scanned images, then tested them with real client files. 4. Day **4** ### Client communication automation Managing Partner, Office Manager On Zoom, the managing partner and office manager set up Copilot to draft engagement letters, onboarding packs and standard update emails from simple prompts. 5. Day **5** ### Advanced Excel and reporting automations 2 Senior Accountants On Zoom, senior accountants created natural-language queries to generate monthly management reports and first drafts of advisory meeting notes, cutting report preparation time by more than half. 6. Day **6** ### Live team training session Entire team and consultant The entire team practised daily AI habits together on Zoom from their own meeting room, built a mini-automation for weekly progress reporting, and walked through each role's new documentation. 7. Day **7** ### Review, documentation handover and retainer discussion Managing Partner We reviewed metrics from the week on Zoom, handed over custom training videos and standard operating procedures, and agreed on a monthly retainer for ongoing improvements. ![Training day on Zoom: the client's team is at their own workplace on a Zoom call, with two accountants building a reconciliation automation in Copilot on a large screen while colleagues take notes on laptops, and the consultant visible on a video screen.](https://completeaibusiness.com/img/cases/accountancy-2.webp) Training day on Zoom: the client's team is at their own workplace on a Zoom call, with two accountants building a reconciliation automation in Copilot on a large screen while colleagues take notes on laptops, and the consultant visible on a video screen. Training ## A live Zoom session, 5 custom videos and a plan per person ### Live training on Zoom A 2-hour interactive workshop on Zoom with the whole team joining from their own meeting room, practising daily AI tasks and building a weekly report automation together. ### Custom training videos - ▶ Setting up your Copilot for personalised email replies 8 min - ▶ Using Copilot in Excel to categorise and match bank transactions 12 min - ▶ Drafting tax return notes and engagement letters in Word 6 min - ▶ Automating meeting notes and follow-ups in Teams 5 min - ▶ Connecting Copilot to Exact Online exports for monthly reports 10 min ### Online training plan, per role - **Managing Partner:** Prioritising advisory conversations over administrative follow-ups - **Senior Accountant:** Using natural language to query financial data in Excel - **Accountant:** Building and refining reconciliation prompts - **Bookkeeper:** Correcting AI extraction errors efficiently - **Payroll Specialist:** Verifying AI-generated compliance checks - **Office Manager:** Automating repetitive Outlook and Teams tasks - **Administrative Assistant:** Using Copilot in Word for document assembly Automations built ## 4 automations that now run every day ### Bank transaction reconciliation **3.5h** per week Before One accountant spent 8 hours per week manually matching bank CSV lines to Exact Online entries and digging into differences. Now Copilot in Excel suggests matches and flags discrepancies. The accountant reviews and finalises in 3 hours per week. ### Client onboarding document pack **3h** per week Before Office manager and admin spent 6 hours per week copying, pasting and personalising engagement letters, privacy forms and welcome emails. Now One prompt with client name and service type generates the full pack in Word and Outlook. Review takes 2 hours per week total. ### Scanned receipt data extraction **3h** per week Before Bookkeepers typed vendor, amount and date from each scanned receipt into Excel, taking 5 hours per week. Now Drag an image into an Excel sheet and Copilot extracts the fields. Bookkeepers verify and correct in 1 hour per week. ### Weekly team report compilation **1.5h** per week Before The administrative assistant spent 3 hours per week gathering updates from email, Teams chats and meetings to create a status report. Now Copilot in Teams pulls together a summary from conversations and shared files. The assistant reviews and sends it in 1 hour. The result ## 45 hours a week back, worth $113,900 a year 34 hours come from everyone using their own AI every day, 11 from the automations. The team now handles 3 more clients without overtime, and error rates on initial bookkeeping entries dropped noticeably. The managing partner spends Fridays working from home for the first time in years. ### Where the hours come from, per week Accountant 4 × 3h daily AI 12h Bookkeeper 4 × 2h daily AI 9h Senior Accountant 2 × 3h daily AI 6h Bank transaction reconciliation automation 3.5h Client onboarding document pack automation 3h Scanned receipt data extraction automation 3h Managing Partner 1 × 2h daily AI 2h Everyone else (4 more) daily AI and smaller automations 6.5h ### Paid back in week 4 At $55 an hour, the time saved earns back the $7,700 in 4 weeks. Every week after that is profit. Week 1 Week 13 Week 26 2,070h back per year $113,900 value per year Retainer: monthly. Monthly retainer covers Copilot license updates, ongoing prompt tuning, and building a new automation for quarterly VAT return preparation. > I was sceptical, but within a week my team stopped dreading the monthly bank reconciliation. We now get it done in under two hours instead of two days. Complete AI Business really understood our specific accounting workflows. **Jan van der Berg**, Managing Partner, Van der Berg Accountants. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "Juniper & Hay: 38 hours a week back with AI" description: "How a 12-person branding and content agency in Manchester, United Kingdom got 38 hours a week back in 7 days with ChatGPT Business: the team, the advice,…" url: https://completeaibusiness.com/case-studies/creative-agency --- ![A normal day at Juniper & Hay: designers at work, client decks on screen, samples and moodboards on the wall.](https://completeaibusiness.com/img/cases/creative-agency-1.webp) Marketing, Advertising and Creative Agencies · Manchester, United Kingdom # Juniper & Hay got 38 hours a week back in 7 days Juniper & Hay is a 12-person branding and content agency in Manchester. They help consumer and B2B brands name, position, design and produce campaigns across web, social and packaging. Most clients are UK and European businesses that need creative work done quickly without losing craft. A normal day at Juniper & Hay: designers at work, client decks on screen, samples and moodboards on the wall. **12** people on the team **38h** back every week **$87,400** value per year at $50/h **5 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** Juniper & Hay and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Brand identity and content for ambitious consumer and B2B brands - **Customers:** D2C food and drink brands, independent retailers, SaaS startups and one national hospitality group. - **Team:** 12 people - **Location:** Manchester, United Kingdom Tools before we started Adobe Creative Cloud Google Workspace Slack Asana Notion Figma HubSpot ### What they sell, share of revenue Brand identity 45% Content campaigns 25% Web and motion 15% Ongoing creative retainer 15% The team ## 12 people, and what their AI now does for them Creative Director Custom GPT for brief critique, brand territory exploration and first-pass creative feedback on design principles. 1 Brand Strategist Research assistant for audience interviews, positioning summaries and competitor brand audits. 1 Senior Designer AI image and layout explorations for moodboards, plus a custom GPT that turns client feedback into design questions. 2 Midweight Designer Image editing support, mockup generation and asset resizing prompts in Canva and Photoshop workflows. 2 Copywriter Brand voice GPT with approved tone examples, plus headline and social post drafting from briefs. 2 Account Manager Meeting notes, project status summaries and client email drafts pulled from Asana and Slack. 2 Studio and Operations Manager Resource planning, timesheet summaries and invoice follow-up templates. 1 New Business Manager Lead research, proposal first drafts and outreach personalization from LinkedIn and HubSpot. 1 The questionnaire ## What Rachel Mercer told us before we started Rachel Mercer, Managing Director, filled in the same questionnaire you can fill in today. Biggest time drain Writing proposals and scope documents from scratch, then turning campaign work into social cutdowns. The goal Save at least 20 hours a week on admin and production so the team can spend more time on original ideas and client relationships. Worries about AI AI will make client work feel generic · Junior team might skip learning core craft · Client data could leak into public tools ### Keep human - Client workshops and brand strategy - Design crits where work gets sharper - Presenting creative work in person ### Take it off our plate - Formatting and re-writing proposal decks - Hunting for project files across Asana and Slack - Turning one campaign into ten social assets - Chasing clients for feedback Our advice ## ChatGPT Business, and the three things to automate first The team already runs most work in Slack, Asana, Notion and Figma, so ChatGPT Business fits without forcing a move to Microsoft or Google. It lets them share custom GPTs for brand voice, proposals and content repurposing, with clear business data controls. Your tools today Adobe Creative Cloud Google Workspace Slack Asana Notion Figma HubSpot We advise **ChatGPT Business** Fit 9/10 Set up for 12 people, connected to your tools AI takes over - Proposal First Draft Generator - Campaign Cutdown Machine - Weekly Client Status Reporter - Design Asset Brief Expander Stays human, AI only supports Creative direction and final craft Client relationships and negotiation Brand strategy judgment calls ### How each platform fit this business ChatGPT Business Chosen 9/10 Claude Team 7/10 Google Gemini for Workspace 6/10 Microsoft 365 Copilot 4/10 ### What we advised to automate first - Proposal and scope first drafts - Campaign content repurposing for social - Project status reports and meeting notes Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### AI Readiness Audit and Platform Setup Rachel Mercer, Operations Manager, IT and Studio Manager Review the questionnaire, map time drains, and set up ChatGPT Business with workspace roles and privacy controls in a Zoom meeting. We agree which workflows to automate first. 2. Day **2** ### Brand Voice and Proposal Assistant Rachel Mercer, New Business Manager, Copywriter Build a custom GPT trained on past winning proposals, services, pricing bands and tone in a Zoom meeting. Start drafting a real proposal from the pipeline. 3. Day **3** ### Content Repurposing Workflow Copywriters, Account Manager Create project folders and custom instructions for turning long-form campaign copy into LinkedIn, Instagram and email cutdowns on Zoom. Test on a live client project. 4. Day **4** ### Design and Image Support Creative Director, Senior Designers, Midweight Designers Set up AI prompts and workflow for moodboard exploration, mockup variations and resizing specs on Zoom. Train designers on how to brief the AI without losing art direction. 5. Day **5** ### Project Reporting and Meeting Notes Account Managers, Operations Manager Connect ChatGPT Business to Asana and Slack in a Zoom meeting, then build a weekly status report prompt that pulls open tasks and team updates into a client-ready summary. 6. Day **6** ### Team Training and Custom Videos Whole team Run individual Zoom setup sessions with screen sharing for all 12 staff, record custom training videos, and make sure each person has their own projects and custom instructions. 7. Day **7** ### Automation Handover and Live Session Whole team Deliver the main automations, run a live Zoom training session with the team in their own meeting room, and agree the monthly maintenance plan. Everyone leaves with documentation and access to videos. ![Rachel and two designers watch on a Zoom call as the consultant walks through the new brand voice GPT.](https://completeaibusiness.com/img/cases/creative-agency-2.webp) Rachel and two designers watch on a Zoom call as the consultant walks through the new brand voice GPT. Training ## A live Zoom session, 6 custom videos and a plan per person ### Live training on Zoom A 45-minute live Zoom session on the final day covered real examples from their own projects, common mistakes, and how to prompt for better creative work. ### Custom training videos - ▶ Using ChatGPT for proposals and scopes 12 min - ▶ Editing AI copy in your brand voice 9 min - ▶ Visual AI prompts for designers 14 min - ▶ Project reporting with Asana and Slack 8 min - ▶ Keeping client data safe in ChatGPT Business 6 min - ▶ Personal AI setup for every role 10 min ### Online training plan, per role - **Creative Director:** Prompting for creative direction and brief writing - **Brand Strategist:** Using AI for research summaries and positioning drafts - **Senior Designer:** Art direction with AI and visual prompt writing - **Midweight Designer:** Faster asset production with AI tools - **Copywriter:** Editing AI drafts in brand voice - **Account Manager:** Clear client update writing and task tracking - **Studio and Operations Manager:** Operations admin and reporting - **New Business Manager:** Proposal drafting and lead research Automations built ## 4 automations that now run every day ### Proposal First Draft Generator **6h** per week Before New business manager spent 6 hours per proposal searching past decks and writing from scratch. Now Pulls structure, pricing, case studies and tone into a draft in 30 minutes. ### Campaign Cutdown Machine **4.5h** per week Before Copywriters manually reworked each campaign into social posts and email snippets. Now One prompt turns approved long-form copy into platform-specific drafts. ### Weekly Client Status Reporter **3h** per week Before Account managers gathered updates from Asana, Slack and email every Friday. Now A weekly report prompt compiles task status and team notes into a client-ready summary. ### Design Asset Brief Expander **2h** per week Before Designers asked the same questions about file formats, sizes and usage every time. Now A custom GPT turns short client requests into clear production specs. The result ## 38 hours a week back, worth $87,400 a year 23 hours come from everyone using their own AI every day, 16 from the automations. The team is saving about 21 hours a week on proposals, cutdowns and reporting, and using that time on client workshops and design craft. Rachel says the agency has not replaced any creative judgment, but the first drafts are far less painful. ### Where the hours come from, per week Proposal First Draft Generator automation 6h Copywriter 2 × 2.5h daily AI 5h Midweight Designer 2 × 2h daily AI 4.5h Campaign Cutdown Machine automation 4.5h Senior Designer 2 × 2h daily AI 3.5h Account Manager 2 × 1.5h daily AI 3h Weekly Client Status Reporter automation 3h Everyone else (5 more) daily AI and smaller automations 8.5h ### Paid back in week 5 At $50 an hour, the time saved earns back the $7,700 in 5 weeks. Every week after that is profit. Week 1 Week 13 Week 26 1,748h back per year $87,400 value per year Retainer: monthly. Covers monthly maintenance, model updates, one personal training hour, and new automations as client services change. > We used to lose Fridays to report formatting and cutdowns. Now a first draft is ready before the team opens the laptop, and the work still sounds like us. **Rachel Mercer**, Managing Director, Juniper & Hay, Juniper & Hay. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "Sunset Ridge Dental: 49 hours a week back with AI" description: "How an 18-person family dental practice with two chairs per dentist in Austin, Texas, USA got 49 hours a week back in 7 days with Microsoft 365 Copilot: the…" url: https://completeaibusiness.com/case-studies/dental-clinic --- ![A calm, sunlit dental office in Austin: a dentist with a patient in the chair while a dental assistant smiles nearby, and in the background, the front desk team talks with a family at check-in. The Eaglesoft screen shows the day's schedule, and a Weave phone sits on the counter.](https://completeaibusiness.com/img/cases/dental-clinic-1.webp) Healthcare Providers and Clinics · Austin, Texas, USA # Sunset Ridge Dental got 49 hours a week back in 7 days We are a family-run practice with two chairs per dentist, offering preventive, restorative, and cosmetic dentistry. Our team of 18 combines experience with gentle care, and we treat patients from toddlers to seniors. We pride ourselves on a personal touch and patient education. A calm, sunlit dental office in Austin: a dentist with a patient in the chair while a dental assistant smiles nearby, and in the background, the front desk team talks with a family at check-in. The Eaglesoft screen shows the day's schedule, and a Weave phone sits on the counter. **18** people on the team **49h** back every week **$94,700** value per year at $42/h **4 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** Sunset Ridge Dental and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Generations of healthy smiles for Austin families since 1995. - **Customers:** Local Austin families, professionals, retirees, and children from neighborhoods like West Lake Hills and Rollingwood. - **Team:** 18 people - **Location:** Austin, Texas, USA Tools before we started Eaglesoft (dental practice management) Microsoft Outlook Microsoft Excel QuickBooks Online Weave (patient communication and phones) Google Drive RingCentral ### What they sell, share of revenue Cleanings & Exams 30% Fillings & Crowns 25% Cosmetic Dentistry (whitening, veneers) 20% Orthodontics (Invisalign) 15% Emergency Care 10% The team ## 18 people, and what their AI now does for them Lead Dentist AI drafts post-visit summaries and patient education handouts based on treatment notes. 2 Associate Dentist AI pre-fills perio charts and generates aftercare instructions from exam findings. 2 Dental Hygienist AI suggests at-home care instructions based on probe readings and patient history. 2 Dental Assistant AI generates supply order lists from consumption logs and patient volume forecasts. 4 Front Desk AI drafts replies to common patient questions about insurance, appointments, and billing. 3 Billing/Insurance Coordinator AI compiles denial reasons and suggests appeal letter templates. 2 Office Manager AI creates weekly reports on schedule utilization and production from raw Eaglesoft data. 1 Marketing Coordinator AI drafts social media posts, patient newsletters, and reviews responses. 1 Sterilization Technician AI manages daily checklist reminders and tracks instrument maintenance logs. 1 The questionnaire ## What Dr. Laura Chen told us before we started Dr. Laura Chen, Co-owner, Lead Dentist, filled in the same questionnaire you can fill in today. Biggest time drain Writing referral letters and explaining treatment plans multiple times. The goal Cut administrative busywork so we can spend more time with patients and leave work on time. Worries about AI Will AI sound robotic with our patients? · Is patient data safe with all these new tools? · Will our team resist learning another system? ### Keep human - Building trust through face-to-face conversations - Diagnosing and treatment planning complex cases - Seeing long-term patient relationships grow ### Take it off our plate - Writing the same explanation for a crown procedure 15 times a day - Manually copying patient data from Eaglesoft into Excel for monthly reports - Handling insurance predetermination paperwork - Chasing lab cases with phone tag Our advice ## Microsoft 365 Copilot, and the three things to automate first Your practice already runs on Outlook, Word, and Excel, so Copilot works directly where your staff already lives. It also meets HIPAA requirements with Microsoft's Business Associate Agreement, and we configured it to respect your patient data boundaries. Your tools today Eaglesoft (dental practice management) Microsoft Outlook Microsoft Excel QuickBooks Online Weave (patient communication and phones) Google Drive RingCentral We advise **Microsoft 365 Copilot** Fit 9/10 Set up for 18 people, connected to your tools AI takes over - Auto-reply drafts for common patient emails - Insurance predetermination letter generator - Monthly production report auto-generator - Patient education handout builder Stays human, AI only supports Clinical diagnosis and treatment planning decisions Sensitive patient conversations and care handoffs Final review of insurance appeals and financial arrangements ### How each platform fit this business Microsoft 365 Copilot Chosen 9/10 ChatGPT Business 7/10 Claude Team 6/10 Google Gemini for Workspace 5/10 ### What we advised to automate first - Patient FAQs and appointment request replies - Insurance predetermination letter drafts - Monthly production and schedule utilization reporting Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### AI Audit and Team Kickoff All 18 team members (morning huddle extended) In a Zoom meeting, we mapped current workflows, identified biggest time drains using the questionnaire, and introduced the team to how AI will help. The whole team attended. 2. Day **2** ### Setting Up Individual AI Workspaces Small groups by role: dentists, hygienists, front desk, billing, office manager In small-group Zoom calls with screen sharing, each role configured their Microsoft 365 Copilot with custom instructions for our practice, and we integrated it with Outlook, Teams, and Eaglesoft via a secure connector. 3. Day **3** ### First Automation Build: Patient FAQs and Appointment Replies Front desk team, billing, and Dr. Chen In a Zoom meeting, we built an automation that drafts replies to common patient messages in Outlook, using your tone and practice policies, saving hours for the front desk. 4. Day **4** ### Insurance Predetermination and Appeal Templates Billing/insurance coordinators, office manager In a Zoom meeting, we created prompts that pull patient data from Eaglesoft to generate predetermination letters and appeal drafts, cutting the billing team's paperwork time in half. 5. Day **5** ### Monthly Reporting Automation Office manager, marketing coordinator, Dr. Chen In a Zoom meeting, we set up a flow where raw Eaglesoft export gets dropped into a folder, and Copilot in Excel generates the production report and appointment utilization chart automatically. 6. Day **6** ### Training Videos and Live Group Session All 18 team members (split into two groups for live practice) We recorded custom videos for each role and held a live Zoom session where every team member practiced their new AI tasks while we coached them. 7. Day **7** ### Refinement, Documentation, and Go-Live Dr. Chen, office manager, and any staff with final questions In a Zoom meeting, we polished automations, wrote one-page guides for each role, and handed over a maintenance plan. The team started using everything live that day. ![On a Zoom call, Dr. Laura Chen and three team members are at their own workplace, practicing prompts while a consultant from Complete AI Business is visible on a screen showing Microsoft Copilot. Sticky notes with workflow ideas cover the whiteboard.](https://completeaibusiness.com/img/cases/dental-clinic-2.webp) On a Zoom call, Dr. Laura Chen and three team members are at their own workplace, practicing prompts while a consultant from Complete AI Business is visible on a screen showing Microsoft Copilot. Sticky notes with workflow ideas cover the whiteboard. Training ## A live Zoom session, 6 custom videos and a plan per person ### Live training on Zoom We brought everyone together on Zoom from the practice's meeting room and practiced writing prompts together, adjusting the AI's tone to sound like our practice, and role-played patient messages until everyone felt natural. ### Custom training videos - ▶ Dentist's Guide to AI-Generated Clinical Summaries 8 min - ▶ Front Desk: Answering Patient Messages with Copilot 6 min - ▶ Billing Team: Streamlining Predeterminations 7 min - ▶ Office Manager's Monthly Reporting in 5 Minutes 5 min - ▶ Everyone: How to Use AI Without Losing Your Voice 4 min - ▶ Data Security and HIPAA with Copilot 3 min ### Online training plan, per role - **Lead Dentist:** Prompt writing for clinical summaries and referrals. - **Associate Dentist:** Integrating AI notes into Eaglesoft charting workflows. - **Dental Hygienist:** Entering perio data faster with voice-to-AI notes. - **Dental Assistant:** Using AI to check insurance predeterminations quickly. - **Front Desk:** Polishing AI responses to keep our friendly tone. - **Billing/Insurance Coordinator:** Feeding claim denials into AI for faster appeals. - **Office Manager:** Data analysis prompts to spot openings quickly. - **Marketing Coordinator:** Generating dental content ideas and first drafts. - **Sterilization Technician:** Quick voice notes for inventory shortages. Automations built ## 4 automations that now run every day ### Auto-reply drafts for common patient emails **8h** per week Before Front desk typed each reply from scratch, often repeating the same information. Now Outlook suggests a draft based on the email content and our policy; staff only tweaks and sends. ### Insurance predetermination letter generator **6h** per week Before Billing coordinators manually typed letters with patient details from Eaglesoft. Now One-prompt generates a letter with patient info pulled from the PMS; coordinators review and edit. ### Monthly production report auto-generator **3h** per week Before Office manager exported raw data from Eaglesoft to Excel, spent 3 hours building charts. Now Export file dropped in a folder triggers Copilot to create the summary report in under a minute. ### Patient education handout builder **5h** per week Before Dentists and hygienists spent time pulling together after-visit summaries and home care instructions. Now AI draft based on treatment codes and doctor notes; ready in seconds. The result ## 49 hours a week back, worth $94,700 a year 27 hours come from everyone using their own AI every day, 22 from the automations. The front desk closes patient messages in half the time, and billing appeals are submitted 40% faster. Dentists no longer write the same post-op instructions over and over, and the office manager gets her Friday report done before lunch. Team morale is up because the busywork shrank, and patients notice we are more present. ### Where the hours come from, per week Auto-reply drafts for common patient emails automation 8h Front Desk 3 × 2h daily AI 6h Billing/Insurance Coordinator 2 × 3h daily AI 6h Insurance predetermination letter generator automation 6h Patient education handout builder automation 5h Dental Assistant 4 × 1h daily AI 4h Dental Hygienist 2 × 1.5h daily AI 3h Everyone else (6 more) daily AI and smaller automations 10.5h ### Paid back in week 4 At $42 an hour, the time saved earns back the $7,700 in 4 weeks. Every week after that is profit. Week 1 Week 13 Week 26 2,254h back per year $94,700 value per year Retainer: monthly. We opted for the $770 monthly retainer so the team can keep getting new automation ideas implemented, get refresher training, and have our custom Copilot instructions updated as insurance rules and software change. > I used to spend three hours every Friday building that production report. Now I drag a file into a folder, and Copilot gives me the charts and a summary. I actually left early last week and had dinner with my kids. **Rita Morales**, Office Manager, Sunset Ridge Dental. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "Vandermeer & Peeters Advocaten: 30 hours a week back with AI" description: "How a 9-person employment and contract law firm in Antwerp, Belgium got 30 hours a week back in 7 days with Microsoft 365 Copilot: the team, the advice, the…" url: https://completeaibusiness.com/case-studies/law-firm --- ![A normal working day at Vandermeer & Peeters: two lawyers review a contract at a desk while a paralegal organises case files.](https://completeaibusiness.com/img/cases/law-firm-1.webp) Legal Services · Antwerp, Belgium # Vandermeer & Peeters Advocaten got 30 hours a week back in 7 days Vandermeer & Peeters is an Antwerp law firm with nine people advising employers on employment law, dismissals, HR policies and commercial contracts. Most clients are small and mid-sized companies in logistics, healthcare and technology. The firm runs on Microsoft 365 and handles sensitive employee and contract matters daily. A normal working day at Vandermeer & Peeters: two lawyers review a contract at a desk while a paralegal organises case files. **9** people on the team **30h** back every week **$103,500** value per year at $75/h **4 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** Vandermeer & Peeters Advocaten and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Employment and contract law for small and mid-sized businesses - **Customers:** HR managers, owners and finance leads at small and mid-sized companies in Antwerp, mainly logistics, healthcare and technology. - **Team:** 9 people - **Location:** Antwerp, Belgium Tools before we started Microsoft Word Outlook Excel Teams SharePoint Clio DocuSign ### What they sell, share of revenue Employment law advice and disputes 40% Contract drafting and review 30% HR compliance packages 20% Litigation representation 10% The team ## 9 people, and what their AI now does for them Managing Partner Copilot in Outlook and Word handles triage, client updates and first drafts of engagement letters. 1 Senior Employment Lawyer Copilot in Word builds first drafts of employment contracts and advice notes from firm templates. 1 Contract Lawyer Copilot in Teams and Word summarises contract reviews and prepares amendment clauses. 2 Paralegal Copilot in Excel and Word prepares bundles, timelines and first research summaries. 2 Legal Secretary Copilot in Outlook schedules, transcribes voicemail and drafts routine client emails. 1 Office Manager Copilot in SharePoint and Excel handles matter lists, billing prep and policy updates. 1 Junior Associate Copilot in Word and Teams prepares legal research notes, case law summaries and client memos. 1 The questionnaire ## What Els Vandermeer told us before we started Els Vandermeer, Managing Partner, filled in the same questionnaire you can fill in today. Biggest time drain Turning client intake notes into first drafts of advice and contracts. The goal Give every fee earner back one to two hours a day and use AI safely within Microsoft tools. Worries about AI Client confidentiality and GDPR · Lawyers relying on AI without checking · Losing the personal voice in legal writing ### Keep human - Client counselling and negotiation - Nuanced drafting where wording matters - Team mentoring and case strategy ### Take it off our plate - Formatting and re-typing standard clauses - Searching old emails for matter history - Time entry and billing narratives - Scheduling across calendars Our advice ## Microsoft 365 Copilot, and the three things to automate first The firm already spends its day in Word, Outlook, Teams and SharePoint, so Microsoft 365 Copilot shows up in the tools staff already know. Microsoft's data controls and EU data residency support the client confidentiality duties of a Belgian employment and contract law firm. Your tools today Microsoft Word Outlook Excel Teams SharePoint Clio DocuSign We advise **Microsoft 365 Copilot** Fit 9/10 Set up for 9 people, connected to your tools AI takes over - Intake to engagement letter draft - Matter email and meeting summary - Contract clause risk extraction - Time narrative and billing prep Stays human, AI only supports Final legal judgement and signing advice Client negotiation and court representation Sensitive employee dismissal conversations ### How each platform fit this business Microsoft 365 Copilot Chosen 9/10 Claude Team 7/10 ChatGPT Business 6/10 Google Gemini for Workspace 5/10 ### What we advised to automate first - Automated first drafts from intake notes to engagement letters and standard advice memos - Email triage and meeting summaries in Outlook and Teams - Contract clause extraction against the firm's playbook Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### Audit and Copilot setup Els, office manager, IT contact One-on-one Zoom call with screen sharing: went through current files, intake forms and templates. Mapped the firm's Microsoft tenant settings and prepared Copilot for the team. 2. Day **2** ### Prompt and document basics Lawyers and paralegals Zoom training for fee earners on secure prompting, drafting from templates and checking sources. Set up shared Copilot prompt libraries. 3. Day **3** ### Email and meeting workflow All staff On Zoom, configured Outlook triage, meeting summaries and Teams notes. Built a shared matter summary format. 4. Day **4** ### Contract and advice automation Els, senior employment lawyer, contract lawyer On Zoom, built the first automation from intake form to draft engagement letter and standard clauses. Tested on two real matter files. 5. Day **5** ### Playbook and clause review Contract lawyers, paralegal On Zoom, created a clause extraction prompt and risk flag list from existing contracts. Paralegals tested it on supplier and employment contracts. 6. Day **6** ### Live team training and Q&A All nine staff Live team Zoom training and Q&A on daily use, privacy settings and checking AI output. Answered questions and adjusted role setups. 7. Day **7** ### Handover and retainer plan Els, office manager On Zoom, reviewed automations, handed over custom training videos and set the retainer tasks for monthly maintenance. ![Client team at their own workplace on a Zoom call, with the Complete AI Business consultant visible on screen.](https://completeaibusiness.com/img/cases/law-firm-2.webp) Client team at their own workplace on a Zoom call, with the Complete AI Business consultant visible on screen. Training ## A live Zoom session, 6 custom videos and a plan per person ### Live training on Zoom All nine staff joined the live Zoom session to practice prompts, set confidentiality rules and work through two real files together. ### Custom training videos - ▶ Secure prompting in Microsoft 365 Copilot 8 min - ▶ Drafting engagement letters from intake notes 10 min - ▶ Summarising meetings and email threads 7 min - ▶ Contract clause extraction and risk flags 12 min - ▶ Outlook triage and scheduling for secretaries 6 min - ▶ Checking AI output and source rules 5 min ### Online training plan, per role - **Managing Partner:** Inbox and meeting summarisation - **Senior Employment Lawyer:** Legal drafting prompts and tone - **Contract Lawyer:** Clause comparison and redlining - **Paralegal:** Document bundling and fact chronology - **Legal Secretary:** Dictation and email triage - **Office Manager:** Internal policy search and timesheet prep - **Junior Associate:** Research prompts and source checks Automations built ## 4 automations that now run every day ### Intake to engagement letter draft **4h** per week Before Secretary retyped intake details into a Word template and booked a lawyer to review. Now Copilot pulls intake form answers into a structured draft engagement letter with the right fee and scope clauses. ### Matter email and meeting summary **3.5h** per week Before Fee earners searched old email chains and reconstructed client timelines before calls. Now Copilot summarises the thread and meeting notes into a one page matter brief. ### Contract clause risk extraction **2.5h** per week Before Contract lawyers manually compared each clause against the firm's playbook and highlighted gaps. Now Copilot extracts clauses and flags missing or unusual terms against the playbook. ### Time narrative and billing prep **2h** per week Before Office manager chased lawyers for activity notes and typed billing narratives. Now Copilot drafts time narratives from calendar and email activity for review. The result ## 30 hours a week back, worth $103,500 a year 18 hours come from everyone using their own AI every day, 12 from the automations. Fee earners report saving around eight hours a week on drafting and email triage, with more time spent on client counselling. The firm now runs all new employment contracts through the playbook automation and has not needed to add admin staff. ### Where the hours come from, per week Contract Lawyer 2 × 2.5h daily AI 4.5h Paralegal 2 × 2h daily AI 4h Intake to engagement letter draft automation 4h Matter email and meeting summary automation 3.5h Junior Associate 1 × 2.5h daily AI 2.5h Contract clause risk extraction automation 2.5h Senior Employment Lawyer 1 × 2h daily AI 2h Everyone else (4 more) daily AI and smaller automations 6.5h ### Paid back in week 4 At $75 an hour, the time saved earns back the $7,700 in 4 weeks. Every week after that is profit. Week 1 Week 13 Week 26 1,380h back per year $103,500 value per year Retainer: monthly. The monthly retainer covers Copilot updates, new clause prompts, quarterly training for new hires and fixes to the intake and billing automations. > The first draft used to take me most of a morning. Now I open Copilot in Word, feed the intake notes and get a solid draft in minutes. My team still checks every point, but we talk to clients instead of retyping clauses. **Els Vandermeer**, Managing Partner, Vandermeer & Peeters Advocaten. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "LarViva Imobiliária: 36 hours a week back with AI" description: "How an 11-person residential real estate agency with a rentals desk in Lisbon, Portugal got 36 hours a week back in 7 days with Microsoft 365 Copilot: the…" url: https://completeaibusiness.com/case-studies/real-estate --- ![A Tuesday morning at LarViva’s office off Avenida da Liberdade. Three agents are on calls with buyers, while another stands at the window pointing out a neighbourhood to a landlord on speakerphone. Post-it notes and property flyers cover a central table.](https://completeaibusiness.com/img/cases/real-estate-1.webp) Real Estate and Property Management · Lisbon, Portugal # LarViva Imobiliária got 36 hours a week back in 7 days A residential real estate agency and rentals desk based in central Lisbon, serving local families, international buyers and property investors. The team of 11 was spending too much time on repetitive admin and struggled to keep rental enquiries moving quickly enough. They wanted clear AI guidance that fit their existing Microsoft tools without disrupting client relationships. A Tuesday morning at LarViva’s office off Avenida da Liberdade. Three agents are on calls with buyers, while another stands at the window pointing out a neighbourhood to a landlord on speakerphone. Post-it notes and property flyers cover a central table. **11** people on the team **36h** back every week **$66,200** value per year at $40/h **6 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** LarViva Imobiliária and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Helping Lisbon families buy, sell and rent homes since 2008. - **Customers:** Local Portuguese families, expats moving to Lisbon, foreign investors and landlords with rental properties in the greater Lisbon area. - **Team:** 11 people - **Location:** Lisbon, Portugal Tools before we started Microsoft 365 (Outlook, Word, Excel, Teams) WhatsApp Business Idealista (property portal) eGO CRM Google Drive Trello ### What they sell, share of revenue Residential sales 55% Rental management and lettings 30% Property management (maintenance, inspections) 10% Property marketing and photography 3% Valuation reports 2% The team ## 11 people, and what their AI now does for them Managing Director Copilot in Outlook and Teams to summarise team updates, draft client emails and prepare weekly activity reports with a few prompts. 1 Sales Agent Each agent uses Copilot in Word to create property listings from bullet points and in Outlook to turn WhatsApp chats into formal offer letters. 3 Rental Agent Copilot in Excel merges landlord info with portal forms to auto-fill comparison tables, and in Teams it summarises viewing feedback into to-do items. 2 Property Manager Copilot in Outlook turns maintenance requests into structured task briefs and drafts repair quotes from photos and notes sent via Teams. 1 Marketing & Communications Copilot in Word and Excel builds social media posts, flyers and market snapshots from listing data stored in Excel, then drafts email campaigns in Outlook. 1 Viewings & Key Holder Copilot in Teams transcribes spoken viewing notes into a schedule and sends them to the right agent, cutting out manual data entry. 1 Client Care / Office Admin Copilot in Outlook categorises incoming emails and drafts replies to common client questions, then logs interactions into eGO CRM via connected flows. 1 Accounts Copilot in Excel turns monthly commission data into invoice drafts and highlights overdue payments from bank statements. 1 The questionnaire ## What Rita Marques told us before we started Rita Marques, Managing Director, filled in the same questionnaire you can fill in today. Biggest time drain Reading and responding to rental enquiries that come from multiple portals and WhatsApp, then copying data into our CRM and Excel sheets. The goal Cut the admin work around every rental and sale so we can handle 20 percent more properties with the same team, without losing the personal touch. Worries about AI Making client communication feel robotic · Agents relying on AI without understanding the property · Data privacy when letting AI read client emails ### Keep human - Building trust with clients face-to-face - Negotiating deals - Evaluating a property's real potential ### Take it off our plate - Rewriting the same boilerplate emails - Manually updating listings on Idealista - Comparing landlord documents against new regulations - Chasing missing paperwork from tenants Our advice ## Microsoft 365 Copilot, and the three things to automate first Your team already runs on Microsoft 365, so Copilot adds AI directly inside Outlook, Teams, Word and Excel without switching tools. It pulls from your existing emails, files and chats to automate the busywork around listings, enquiries and reports, which is exactly where you spend too much time. Your tools today Microsoft 365 (Outlook, Word, Excel, Teams) WhatsApp Business Idealista (property portal) eGO CRM Google Drive Trello We advise **Microsoft 365 Copilot** Fit 10/10 Set up for 11 people, connected to your tools AI takes over - Rental enquiry to first reply and CRM log - Weekly agent activity report from Teams notes - Maintenance quote brief from photos and messages - Commission invoice draft from Excel deal tracker Stays human, AI only supports Negotiation calls and final price discussions Property valuation judgment and in-person appraisals Client relationship building and emotional reassurance ### How each platform fit this business Microsoft 365 Copilot Chosen 10/10 ChatGPT Business 7/10 Claude Team 6/10 Google Gemini for Workspace 5/10 ### What we advised to automate first - Drafting rental enquiry replies and offer letters - Creating comparison spreadsheets from landlord documents - Summarising Teams viewing notes into agent to-do lists Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### AI Readiness and Platform Setup Managing Director, Office Admin In a Zoom meeting, we activated Microsoft 365 Copilot for the whole team, set governance rules and explained how data stays private. Rita joined on Zoom to approve permissions. 2. Day **2** ### Mapping the Busywork Full team (11 people, rotated in small groups) In a Zoom session, we listed every repetitive task the team hated, grouped them by role and picked the top 10 to automate. Sales and rental agents walked us through a typical enquiry on the call. 3. Day **3** ### Writing the First Automations Sales Agents, Rental Agents, Marketing In a Zoom meeting, we built Copilot prompts for property descriptions, offer letters and enquiry replies. Agents tested them with live cases on the call and gave feedback on tone. 4. Day **4** ### Spreadsheets and Data Work Accounts, Property Manager, Rental Agent In a Zoom session, we showed Accounts and Property Manager how to create templates that auto-fill from emails and teams chats, then built a live comparison sheet for a new landlord. 5. Day **5** ### CRM and Follow-Up Flows Office Admin, Sales Agents, Viewings & Key Holder In a Zoom meeting, we linked Copilot outputs to eGO CRM via quick-capture steps, and created a follow-up checklist that prompts agents at the right time. Admin tested the full loop on the call. 6. Day **6** ### Personal AI Setups and Training Full team (one-on-one Zoom slots) In one-on-one Zoom calls with screen sharing, each team member set up their own Copilot custom instructions and quick-access prompts. We recorded their screens to build the custom video library. 7. Day **7** ### Live Training and Wrap-Up Full team, consultant We ran a one-hour live training on Zoom with the whole team, demonstrated all automations, answered questions and agreed what goes into the retainer. ![At the client's workplace during the AI training session on Zoom, the consultant appears on a screen and shows Ana (sales) and Tiago (rentals) how to turn a WhatsApp voice note into a structured offer email. A laptop is open to Teams, and the team's screen shows the Copilot prompt.](https://completeaibusiness.com/img/cases/real-estate-2.webp) At the client's workplace during the AI training session on Zoom, the consultant appears on a screen and shows Ana (sales) and Tiago (rentals) how to turn a WhatsApp voice note into a structured offer email. A laptop is open to Teams, and the team's screen shows the Copilot prompt. Training ## A live Zoom session, 5 custom videos and a plan per person ### Live training on Zoom We walked the entire team through the three main automations live on Zoom, from enquiry to signed contract, and ran a Q&A on how to keep the tone warm and human. ### Custom training videos - ▶ Turning a WhatsApp message into a client offer letter 8 min - ▶ Building a landlord comparison sheet from emailed documents 11 min - ▶ Generating weekly team summaries from Teams chats 7 min - ▶ Creating a property listing from a one-line description 6 min - ▶ Setting up maintenance task briefs from photos 9 min ### Online training plan, per role - **Managing Director:** Using Copilot in Teams to manage leadership workflows - **Sales Agent:** Property description drafting and email follow-up with Copilot - **Rental Agent:** Quick client updates and viewing summaries with AI - **Property Manager:** Streamlining maintenance reporting and quotes - **Marketing & Communications:** AI-assisted content creation from property data - **Viewings & Key Holder:** Hands-free viewing notes and scheduling - **Client Care / Office Admin:** Email triage and CRM logging shortcuts - **Accounts:** Financial summaries and payments tracking Automations built ## 4 automations that now run every day ### Rental enquiry to first reply and CRM log **4h** per week Before Agent read the portal email or WhatsApp, looked up the property on Idealista, drafted a reply, pasted info into CRM. Took 12 minutes per enquiry. Now Copilot pulls property details from a shared listing spreadsheet, drafts a personalised reply and generates a CRM entry that the agent copies in with one click. Average 3 minutes per enquiry. ### Weekly agent activity report from Teams notes **3h** per week Before Sales and rental agents sent bullet-point updates to Rita, who spent an hour compiling them into a weekly summary for the Monday meeting. Now Copilot gathers channel posts and 1:1 chat summaries, then drafts the weekly report inside Teams. Rita reviews and sends it in 15 minutes. ### Maintenance quote brief from photos and messages **3h** per week Before Property Manager got WhatsApp photos and voice notes, then manually wrote a task brief and emailed it to contractors, back and forth for details. Now Copilot interprets photos and texts in Teams, creates a structured brief with room, issue and urgency, and drafts an email to contractors. Brief ready in 4 minutes. ### Commission invoice draft from Excel deal tracker **3h** per week Before Accounts pulled numbers from the deal tracker, cross-checked with signed contracts, wrote invoice drafts by hand and sent them for approval. Now Copilot reads the Excel tracker, matches fields to an invoice template and generates a draft in Word, ready for a final check. The result ## 36 hours a week back, worth $66,200 a year 23 hours come from everyone using their own AI every day, 13 from the automations. The team now starts every new enquiry inside Copilot, cutting first-response time from 45 minutes to under 5. Rental agents handle 30 percent more enquiries without extra hires, and the Monday meeting is down to 20 minutes because the report already landed in Rita's inbox. ### Where the hours come from, per week Sales Agent 3 × 2.5h daily AI 7h Rental Agent 2 × 2h daily AI 4h Rental enquiry to first reply and CRM log automation 4h Marketing & Communications 1 × 3h daily AI 3h Weekly agent activity report from Teams notes automation 3h Maintenance quote brief from photos and messages automation 3h Commission invoice draft from Excel deal tracker automation 3h Everyone else (5 more) daily AI and smaller automations 8.5h ### Paid back in week 6 At $40 an hour, the time saved earns back the $7,700 in 6 weeks. Every week after that is profit. Week 1 Week 13 Week 26 1,656h back per year $66,200 value per year Retainer: monthly. The monthly retainer keeps the automated templates updated with regulation changes, adds new Copilot prompts as the rental portfolio grows, and refreshes the training videos when Microsoft releases major updates. > I was worried my team would sound like robots, but the system we built actually sounds more personal because it remembers details I forget. We now answer a rental enquiry before lunch that would have sat until Tuesday. It just works with Outlook and Teams like we always did. **Rita Marques**, Managing Director, LarViva Imobiliária, LarViva Imobiliária. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "EcoCulinar: 25 hours a week back with AI | Complete AI Business" description: "How an 8-person online store for sustainable kitchenware in Copenhagen, Denmark got 25 hours a week back in 7 days with Google Gemini for Workspace: the…" url: https://completeaibusiness.com/case-studies/online-store --- ![A Tuesday morning at EcoCulinar's Copenhagen office: Maja sketches a spoon design at her standing desk while a customer support agent reviews AI-drafted emails on her laptop, sample pots and packaging all around.](https://completeaibusiness.com/img/cases/online-store-1.webp) E-commerce and D2C · Copenhagen, Denmark # EcoCulinar got 25 hours a week back in 7 days EcoCulinar designs and sells plastic-free kitchen tools, cookware, and storage. Based in Copenhagen, we ship directly to customers across Europe. Everything we make is dishwasher safe and built to last. A Tuesday morning at EcoCulinar's Copenhagen office: Maja sketches a spoon design at her standing desk while a customer support agent reviews AI-drafted emails on her laptop, sample pots and packaging all around. **8** people on the team **25h** back every week **$43,700** value per year at $38/h **9 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** EcoCulinar and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Sustainable kitchenware for modern European homes. - **Customers:** Eco-conscious home cooks, mostly aged 25-45, across Europe. - **Team:** 8 people - **Location:** Copenhagen, Denmark Tools before we started Shopify Google Workspace Slack Mailchimp QuickBooks Canva Asana ### What they sell, share of revenue Cookware 40% Kitchen Tools 25% Food Storage 20% Subscriptions 10% Textiles & Accessories 5% The team ## 8 people, and what their AI now does for them Founder & CEO Gemini drafts investor updates, summarises monthly metrics from Sheets, and generates design brief outlines. 1 Operations Manager Gemini monitors inventory Google Sheets, flags low stock, and drafts SOPs from meeting notes. 1 Marketing Lead Gemini writes campaign briefs, analyses Mailchimp performance data, and suggests social media captions. 1 Content Creator Gemini generates initial product descriptions, blog post ideas, and alt-text from product specs in Docs. 1 Customer Support Gemini drafts replies to common order and shipping questions, and summarises ticket histories in Gmail. 2 Logistics Coordinator Gemini pulls shipping data into Sheets, drafts weekly delivery reports, and flags delayed orders. 1 Product Designer Gemini researches sustainable material properties and creates mood board summaries from prompts. 1 The questionnaire ## What Maja Hansen told us before we started Maja Hansen, Founder & CEO, filled in the same questionnaire you can fill in today. Biggest time drain Managing customer inquiries and order updates manually. The goal Free up 20 hours a week so the team can focus on growth, not admin. Worries about AI AI might give incorrect product information to customers · Team could resist change · Losing the personal touch in customer service ### Keep human - Creating new product designs - Talking to suppliers about sustainable materials - Writing the weekly newsletter ### Take it off our plate - Answering repetitive customer emails - Updating inventory spreadsheets - Scheduling social media posts - Generating product descriptions for hundreds of SKUs Our advice ## Google Gemini for Workspace, and the three things to automate first It sits directly inside the Google Workspace tools your team already uses every day, so there is no new app to learn. The business-grade version keeps your data private and gives every employee an AI assistant in Gmail, Docs, and Sheets. Your tools today Shopify Google Workspace Slack Mailchimp QuickBooks Canva Asana We advise **Google Gemini for Workspace** Fit 9/10 Set up for 8 people, connected to your tools AI takes over - Customer inquiry triage and draft reply - Product description generator - Weekly logistics and sales report Stays human, AI only supports Handling complex customer complaints Creative product design decisions Supplier negotiations ### How each platform fit this business Google Gemini for Workspace Chosen 9/10 ChatGPT Business 7/10 Claude Team 6/10 Microsoft 365 Copilot 2/10 ### What we advised to automate first - Customer inquiry classification and response drafting in Gmail - Product description generation from supplier specs in Docs - Order status updates and weekly delivery reports in Sheets Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### Discovery and platform setup Maja (Founder), Operations Manager In a Zoom meeting, we mapped the team's daily workflows and identified the highest-friction tasks. Then we provisioned Gemini for Workspace for all eight accounts and set up the admin controls. 2. Day **2** ### Individual AI configurations Full team Every team member joined a one-on-one Zoom call with screen sharing to configure their own Gemini projects, custom instructions, and connected apps. We built personal prompt libraries for their specific roles. 3. Day **3** ### Daily AI use training Full team In a Zoom training session, we ran hands-on exercises in Gmail, Docs, and Sheets. Each person practised drafting emails, analysing data, and writing in their own voice with Gemini's help. 4. Day **4** ### Automation building part one Customer Support, Operations Manager In a Zoom meeting, we built the customer inquiry triage system in Gmail and connected it to a shared label structure. Support staff tested the draft-and-review flow. 5. Day **5** ### Content and marketing automations Content Creator, Marketing Lead In a Zoom meeting, we set up a product description generator that pulls from supplier datasheets in Docs, and created templates for social media captions and blog outlines. 6. Day **6** ### Reporting and logistics automations Logistics Coordinator, Operations Manager In a Zoom meeting, we automated a weekly sales report that compiles Shopify data into Sheets and generates a plain-English summary. Logistics triggers were added for shipment delays. 7. Day **7** ### Live training and go-live review Full team plus Complete AI Business consultant The whole team joined a live 90-minute Zoom session to practise their new setups, ask final questions, and document everyday AI habits. We closed with a plan for the first 30 days. ![The Complete AI Business consultant leads the live Day 7 workshop on a Zoom call. The client's team is at their own workplace, with the consultant visible on a screen showing Gemini's interface, while three employees take notes and one tests a prompt on her phone.](https://completeaibusiness.com/img/cases/online-store-2.webp) The Complete AI Business consultant leads the live Day 7 workshop on a Zoom call. The client's team is at their own workplace, with the consultant visible on a screen showing Gemini's interface, while three employees take notes and one tests a prompt on her phone. Training ## A live Zoom session, 5 custom videos and a plan per person ### Live training on Zoom A 90-minute Zoom workshop where every team member practised with their own Gemini setup, asked questions live, and we role-played common customer scenarios. ### Custom training videos - ▶ Getting started with Gemini in Gmail 6 min - ▶ Drafting customer replies with your voice 8 min - ▶ Product description creation in Docs 5 min - ▶ Weekly sales reports with Sheets and Gemini 7 min - ▶ Prompt writing for brand-consistent content 4 min ### Online training plan, per role - **Founder & CEO:** Strategic planning and market analysis with Gemini. - **Operations Manager:** Automating operations routines with Google Sheets and Gemini. - **Marketing Lead:** Campaign planning and data insights with AI. - **Content Creator:** Writing faster without losing brand voice. - **Customer Support:** Handling repetitive emails efficiently with AI drafts. - **Logistics Coordinator:** Automating shipping reports and supplier emails. - **Product Designer:** Material research and design inspiration with AI. Automations built ## 3 automations that now run every day ### Customer inquiry triage and draft reply **7h** per week Before Support agents read and wrote every reply from scratch for 50+ emails a day. Now Gemini classifies incoming emails and drafts a reply; the agent only reviews and sends. ### Product description generator **4.5h** per week Before The content creator wrote each new product description manually from supplier specs. Now Gemini drafts a description based on a formatted spec sheet in Docs; the creator polishes the copy. ### Weekly logistics and sales report **2h** per week Before The logistics coordinator pulled data from Shopify manually and typed a summary every Friday. Now A scheduled pull fills a Google Sheet, and Gemini writes a summary with trends and alerts. The result ## 25 hours a week back, worth $43,700 a year 12 hours come from everyone using their own AI every day, 14 from the automations. The team now handles 40% more customer emails in the same time, and new product listings go live three days faster. Maja has reclaimed her Friday afternoons for design work, and nobody dreads the weekly report anymore. ### Where the hours come from, per week Customer inquiry triage and draft reply automation 7h Product description generator automation 4.5h Customer Support 2 × 2h daily AI 3.5h Marketing Lead 1 × 2h daily AI 2h Content Creator 1 × 2h daily AI 2h Weekly logistics and sales report automation 2h Operations Manager 1 × 1.5h daily AI 1.5h Everyone else (3 more) daily AI and smaller automations 2.5h ### Paid back in week 9 At $38 an hour, the time saved earns back the $7,700 in 9 weeks. Every week after that is profit. Week 1 Week 13 Week 26 1,150h back per year $43,700 value per year Retainer: monthly. We chose a monthly retainer so we get ongoing priority support, model updates, and new automations for things like returns processing and personalised product recommendations. > Before, our support team was drowning in 'where is my order?' emails. Now Gemini drafts a perfect reply in seconds, and we just add a quick personal note. I finally had time last month to design a new bamboo utensil line instead of spending my evenings on admin. **Maja Hansen**, Founder & CEO, EcoCulinar. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "De Vries Renovatie & Installatie: 45 hours a week back with AI" description: "How a 21-person renovation and installation contractor (office team plus field crews) in Rotterdam, the Netherlands got 45 hours a week back in 7 days with…" url: https://completeaibusiness.com/case-studies/renovation --- ![Morning huddle at the Rotterdam office: project managers align with field crews before heading to sites.](https://completeaibusiness.com/img/cases/renovation-1.webp) Construction and Engineering · Rotterdam, the Netherlands # De Vries Renovatie & Installatie got 45 hours a week back in 7 days A family-run contractor with 21 people, completing around 90 projects a year. They handle everything from kitchen refits to full commercial plumbing and electrical installations. The office team manages estimating, scheduling, and invoicing while field crews execute the work across the city. Morning huddle at the Rotterdam office: project managers align with field crews before heading to sites. **21** people on the team **45h** back every week **$93,200** value per year at $45/h **4 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** De Vries Renovatie & Installatie and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Rotterdam renovation and installation contractor for homes and businesses - **Customers:** Homeowners, property managers, and small business owners in and around Rotterdam. - **Team:** 21 people - **Location:** Rotterdam, the Netherlands Tools before we started Microsoft 365 (Outlook, Excel, Teams) Exact Online WhatsApp Google Maps Adobe Acrobat Trello ### What they sell, share of revenue Residential Renovations 40% Commercial Installations 30% Maintenance Contracts 20% Emergency Repair Services 10% The team ## 21 people, and what their AI now does for them Director Copilot in Teams and Outlook to summarize project updates and draft client communications. 1 Project Manager AI-assisted daily log creation and schedule adjustments in Excel and Teams. 2 Estimator Copilot in Excel to pull material costs from past jobs and generate quote drafts. 2 Office Manager Automated document filing in SharePoint and email responses for common inquiries. 1 Finance & HR Officer AI to cross-check invoices against contracts and draft payment reminders. 1 Site Supervisor Voice-to-text in OneNote for site notes and automatic report creation. 2 Field Technician AI chatbot in Teams for quick access to installation manuals and safety guidelines. 12 The questionnaire ## What Daan de Vries told us before we started Daan de Vries, Director, filled in the same questionnaire you can fill in today. Biggest time drain Manual data entry across estimating, scheduling, and invoicing. The goal Cut admin time by 40% so we can take on two extra projects per month without hiring office staff. Worries about AI Data privacy of client information · Field workers not adopting new tools · AI misunderstanding complex project specs ### Keep human - Client consultations - On-site problem solving - Quality control walkthroughs ### Take it off our plate - Updating Excel spreadsheets - Chasing late invoices - Translating scribbled field notes into formal reports - Scheduling crew across multiple projects Our advice ## Microsoft 365 Copilot, and the three things to automate first Your company runs entirely on Microsoft 365 for documents, spreadsheets, and communication. Copilot plugs directly into that and can use your existing data, so you start getting help from day one without migrating anything. Your tools today Microsoft 365 (Outlook, Excel, Teams) Exact Online WhatsApp Google Maps Adobe Acrobat Trello We advise **Microsoft 365 Copilot** Fit 9/10 Set up for 21 people, connected to your tools AI takes over - Quote Generator - Voice-to-Report Site Logs - Payment Reminder Workflow - Crew Schedule Optimizer Stays human, AI only supports Client negotiations and relationship building On-site quality inspections Crew training and safety briefings ### How each platform fit this business Microsoft 365 Copilot Chosen 9/10 ChatGPT Business 7/10 Claude Team 6/10 Google Gemini for Workspace 5/10 ### What we advised to automate first - Estimate and quote generation - Site report compilation from voice notes - Invoice follow-up reminders Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### Audit & AI Roadmap Director, office manager, lead estimator In a Zoom meeting, we mapped your current workflows for estimating, project reporting, and invoicing. Identified the top time-wasting tasks and set goals for the week. 2. Day **2** ### Copilot Setup & Data Connection Director, office manager In a Zoom call with screen sharing, configured Microsoft 365 Copilot with access to your SharePoint files, Outlook, and Excel templates. Set up security groups for different roles. 3. Day **3** ### Role-by-Role AI Configuration Project managers, estimators, finance & HR officer In a Zoom meeting, we built custom prompts and connected apps for each department: PM dashboards, estimator Excel models, and finance invoice alerts. 4. Day **4** ### Automation Building Sprint Estimators, project managers, office manager In a Zoom meeting, created the three core automations: estimate drafting, site report generator, and invoice reminder flow. Tested with real data. 5. Day **5** ### Mobile AI for Field Crews Site supervisors, 4 field technicians In a Zoom meeting, showed site supervisors and field techs how to use voice notes in Teams and access safety docs on their phones. Set up quick templates. 6. Day **6** ### All-Hands Training & Practice Entire team Live Zoom training session with the whole team in their own meeting room, doing hands-on exercises: creating a quote, turning a voice memo into a report, and sending an automated invoice reminder. 7. Day **7** ### Handover & Documentation Director, office manager In a Zoom meeting, delivered custom training videos, written guides, and a maintenance plan. Reviewed next steps and set up a monthly check-in. ![Consultant leads the team through voice-to-report training on Zoom, with field techs testing the mobile app at their own workplace.](https://completeaibusiness.com/img/cases/renovation-2.webp) Consultant leads the team through voice-to-report training on Zoom, with field techs testing the mobile app at their own workplace. Training ## A live Zoom session, 6 custom videos and a plan per person ### Live training on Zoom A 90-minute hands-on Zoom workshop where every team member practiced their new daily AI routine, from field technicians recording voice notes to estimators drafting quotes. ### Custom training videos - ▶ How to Generate a Quote from an Email 9 min - ▶ Turning Voice Notes into Daily Reports 6 min - ▶ Automating Overdue Invoice Reminders 5 min - ▶ Quick Site Safety Checklist with AI 4 min - ▶ Using Copilot in Teams for Project Updates 8 min - ▶ Excel Estimating with AI Assistance 10 min ### Online training plan, per role - **Director:** Executive summaries and email delegation - **Project Manager:** Creating automated status reports - **Estimator:** Using AI for rapid cost lookups - **Office Manager:** Organizing digital files with Copilot - **Finance & HR Officer:** Invoice reconciliation with AI - **Site Supervisor:** Voice reporting and photo notes - **Field Technician:** Asking installation questions via Teams Automations built ## 4 automations that now run every day ### Quote Generator **9h** per week Before Estimators manually searched past projects and typed quotes in Excel, spending about 4 hours per quote. Now Copilot drafts a quote in Excel from the client email, pulling material prices and labour rates; estimator reviews and tweaks in under 1 hour. ### Voice-to-Report Site Logs **3h** per week Before Site supervisors spent a total of 4 hours per week typing formal reports from paper notes. Now They speak their notes, AI creates a draft report, trimming total report time to 1 hour per week. ### Payment Reminder Workflow **2.5h** per week Before Office manager manually tracked overdue invoices in Exact Online and emailed reminders, 3 hours per week. Now Automated flow scans overdue invoices daily and sends a reminder email draft for the manager to approve in 30 minutes. ### Crew Schedule Optimizer **4h** per week Before Project managers spent 5 hours per week moving crew members around in Excel based on project changes. Now Copilot suggests schedule updates in Teams based on project timelines, PM reviews and confirms in 1 hour. The result ## 45 hours a week back, worth $93,200 a year 27 hours come from everyone using their own AI every day, 19 from the automations. Project turnaround time dropped by 15% because estimates go out faster. The office manager now handles invoicing without overtime, and site supervisors submit reports on the same day. ### Where the hours come from, per week Quote Generator automation 9h Estimator 2 × 3.5h daily AI 7h Project Manager 2 × 3h daily AI 6h Field Technician 12 × 0.5h daily AI 6h Crew Schedule Optimizer automation 4h Voice-to-Report Site Logs automation 3h Payment Reminder Workflow automation 2.5h Everyone else (4 more) daily AI and smaller automations 7.5h ### Paid back in week 4 At $45 an hour, the time saved earns back the $7,700 in 4 weeks. Every week after that is profit. Week 1 Week 13 Week 26 2,070h back per year $93,200 value per year Retainer: monthly. Monthly retainer covers Copilot license management, bi-monthly virtual check-ins, and two new automation tweaks per month based on user feedback. > The AI took the paperwork off my desk. I used to spend Fridays cleaning up invoices and schedules; now that happens automatically. The field guys actually enjoy the voice reports because they hate typing. We've taken on two extra renovation projects since going through the programme, and I'm not working longer hours. **Daan de Vries**, Director, De Vries Renovatie & Installatie. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "The Thistle & Thorn: 37 hours a week back with AI" description: "How a 19-person 34-room boutique hotel with a restaurant in Edinburgh, United Kingdom got 37 hours a week back in 7 days with Microsoft 365 Copilot: the…" url: https://completeaibusiness.com/case-studies/boutique-hotel --- ![A Monday morning at The Thistle & Thorn: a front desk agent greets a guest while a Copilot-drafted email sits ready for review on the second screen. In the kitchen, chefs work from AI-generated prep cards.](https://completeaibusiness.com/img/cases/boutique-hotel-1.webp) Hospitality, Hotels and Restaurants · Edinburgh, United Kingdom # The Thistle & Thorn got 37 hours a week back in 7 days The Thistle & Thorn is an independent hotel set in a listed building, blending contemporary design with the city's history. Its restaurant, Bramble, draws locals and guests for dishes built around small-scale producers. The close-knit team of 19 runs every part of the guest experience, from housekeeping to event hosting. A Monday morning at The Thistle & Thorn: a front desk agent greets a guest while a Copilot-drafted email sits ready for review on the second screen. In the kitchen, chefs work from AI-generated prep cards. **19** people on the team **37h** back every week **$54,500** value per year at $32/h **7 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** The Thistle & Thorn and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## 34-room boutique hotel with a seasonal Scottish restaurant in Edinburgh's Old Town - **Customers:** Leisure tourists, business travellers, and Edinburgh locals looking for a relaxed, authentic place to eat and stay. - **Team:** 19 people - **Location:** Edinburgh, United Kingdom Tools before we started Mews (property management) Square for Restaurants (POS) Xero (accounting) Microsoft 365 (Outlook, Word, Excel, Teams) Mailchimp (email marketing) Trello (task management) SiteMinder (channel manager) ### What they sell, share of revenue Hotel rooms 65% Restaurant dining 25% Events and private dining 7% Bar and lounge 3% The team ## 19 people, and what their AI now does for them General Manager Copilot drafts board reports, summarises guest feedback, and generates financial forecasts in Excel. 1 Front Desk Manager AI handles guest email drafts, automated booking confirmations, and shift summary notes in Teams. 1 Front Desk Agents AI drafts responses to the ten most common guest questions and pulls up reservation details for personalised replies. 4 Housekeeping Manager Copilot creates daily room assignment lists based on Mews occupancy data and flags VIP preferences. 1 Housekeepers Receive AI-generated personal cleaning lists each morning via Teams, noting guest requests and early check-ins. 5 Head Chef AI suggests daily specials based on stock levels, seasonality, and cost-per-plate calculations. 1 Sous Chef AI monitors inventory and sends Teams alerts when items drop below the level needed for upcoming covers. 1 Line Cooks Get prep lists and scaled recipes from AI based on the day's reservations and event orders. 2 Restaurant Manager AI drafts the weekly staff rota, handles event enquiry emails, and builds the pre-shift briefing note. 1 Servers AI suggests dish descriptions and wine pairings on a tablet, and captures dietary notes into a shareable log. 2 The questionnaire ## What Fiona MacLeod told us before we started Fiona MacLeod, General Manager, filled in the same questionnaire you can fill in today. Biggest time drain Handling repetitive guest email inquiries and manually compiling daily reports. The goal Free up 15 hours per week across the team so we can spend more time on guest experience and creative projects. Worries about AI Staff might struggle with new technology · Losing the personal touch with guests · Data security with AI ### Keep human - Welcoming guests personally - Designing seasonal menus - Training new staff ### Take it off our plate - Chasing suppliers for orders - Entering data into spreadsheets - Writing the same email for the tenth time - Creating staff rotas Our advice ## Microsoft 365 Copilot, and the three things to automate first Your team already lives in Outlook, Word, Excel, and Teams every day. Copilot works right inside those apps, so nobody has to learn a new tool, and it draws on your existing emails and documents to give relevant, contextual help. Your tools today Mews (property management) Square for Restaurants (POS) Xero (accounting) Microsoft 365 (Outlook, Word, Excel, Teams) Mailchimp (email marketing) Trello (task management) SiteMinder (channel manager) We advise **Microsoft 365 Copilot** Fit 9/10 Set up for 19 people, connected to your tools AI takes over - Guest Email Autoresponder - Daily Revenue & Occupancy Snapshot - Weekly Staff Rota Generator - Low Stock Alerts for Kitchen Stays human, AI only supports in-person guest check-ins and conversations menu development and tasting staff recognition and coaching ### How each platform fit this business Microsoft 365 Copilot Chosen 9/10 Google Gemini for Workspace 7/10 ChatGPT Business 6/10 Claude Team 5/10 ### What we advised to automate first - responding to common guest email inquiries - generating daily occupancy and revenue reports - creating weekly staff rotas Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### AI Readiness Audit & Team Brainstorm Fiona MacLeod (GM), Front Desk Manager, Housekeeping Manager, Head Chef, Restaurant Manager We mapped every workflow across front desk, housekeeping, kitchen, and restaurant. Department heads listed 32 repetitive tasks, scored each one for impact, and we picked the top 5 to tackle first. 2. Day **2** ### Copilot Setup & Integration Fiona, IT lead, all department heads Activated Microsoft 365 Copilot for all 19 accounts, linked it to their existing SharePoint folders, Teams channels, and Outlook. Configured guest-data isolation so personal details stay within the hotel's tenancy. 3. Day **3** ### Individual AI Training Plans All 19 staff, in small groups by role Met with each role group on Zoom with screen sharing to design their daily Copilot prompts, custom instructions, and saved phrases. Built a shared prompt library in Teams so everyone uses the hotel's tone. 4. Day **4** ### Building the Guest Email Automation Fiona, Front Desk Manager Created a Copilot flow that reads incoming guest emails, pulls the booking from Mews via a CSV export, and drafts a personalised reply. Front desk now reviews and sends with one click, instead of typing from scratch. 5. Day **5** ### Automating Reports & Rostering Fiona, Restaurant Manager, Head Chef Linked Copilot to Xero and Mews exports so the daily revenue and occupancy report pops into Fiona's inbox each morning. Built a rota prompt that suggests shifts based on forecast covers and staff availability. 6. Day **6** ### Live Team Training Session All 19 staff, plus the Complete AI Business consultant on Zoom A 90-minute hands-on workshop on Zoom. The team could join together from their own meeting room or each from their own desk. Each person built their own Copilot assistant, practised writing a guest welcome note, and the kitchen team tried scaling a recipe with AI. 7. Day **7** ### Documentation, Handover & Retainer Fiona, Front Desk Manager, Housekeeping Manager, Restaurant Manager Delivered custom video guides, a written playbook with all prompts, and a list of 15 future automation ideas. Fiona signed up for a monthly retainer to keep everything running smoothly. ![During the live training session on Zoom, the whole team gathers at their own workplace around laptops and phones, building their own AI assistants while the consultant guides them from a screen.](https://completeaibusiness.com/img/cases/boutique-hotel-2.webp) During the live training session on Zoom, the whole team gathers at their own workplace around laptops and phones, building their own AI assistants while the consultant guides them from a screen. Training ## A live Zoom session, 6 custom videos and a plan per person ### Live training on Zoom A 90-minute Zoom workshop where all 19 team members set up their personal Copilot, practised drafting real guest emails from that day's inbox, and tested the new automations live on the call. ### Custom training videos - ▶ Getting Started with Copilot in Outlook 6 min - ▶ Writing a Perfect Guest Email with AI 5 min - ▶ Automating Your Daily Report in Excel 8 min - ▶ Using Copilot in Teams for Shift Handovers 4 min - ▶ AI-Assisted Menu Planning with Copilot 7 min - ▶ Building Your Own Custom Prompts 9 min ### Online training plan, per role - **General Manager:** Using Copilot for strategic analysis and communication - **Front Desk Manager:** Streamlining front desk email and reporting tasks - **Front Desk Agents:** Quick AI-assisted guest communication - **Housekeeping Manager:** Efficient scheduling with AI prompts - **Housekeepers:** Reading and acting on AI-generated task lists - **Head Chef:** AI-assisted menu planning and cost calculation - **Sous Chef:** Automating supplier and inventory tasks - **Line Cooks:** Using AI for daily prep instructions - **Restaurant Manager:** AI for scheduling and event communication - **Servers:** AI-assisted table service and upselling tips Automations built ## 4 automations that now run every day ### Guest Email Autoresponder **5h** per week Before Front desk manually typed replies to booking changes, early check-in requests, and FAQs, often repeating the same information. Now Copilot drafts a reply using the guest's name, booking details from Mews, and the hotel's style guide. Staff review and send in one click. ### Daily Revenue & Occupancy Snapshot **3h** per week Before Fiona spent 30 minutes each morning exporting data from Mews and Xero, pasting it into an Excel template, and writing commentary. Now A Copilot prompt pulls the latest Mews and Xero exports from SharePoint, creates a one-page summary with charts, and emails it to Fiona by 7:30am. ### Weekly Staff Rota Generator **4h** per week Before Restaurant and housekeeping managers spent hours each week juggling availability, occupancy forecasts, and shift swaps in a shared spreadsheet. Now Managers enter key numbers into a Copilot prompt. It outputs a draft rota in an Excel table, ready for final adjustments and a Teams post. ### Low Stock Alerts for Kitchen **2h** per week Before Sous chef manually checked inventory against upcoming reservations every two days and often discovered shortages at the last minute. Now Copilot watches the shared inventory spreadsheet and daily cover numbers. When a key item drops below the predicted need, it sends a Teams alert to the sous chef. The result ## 37 hours a week back, worth $54,500 a year 23 hours come from everyone using their own AI every day, 14 from the automations. The hotel now saves 22 hours per week, and front desk handles email queries twice as fast without losing a personal tone. The team built three more automations on their own, including a guest feedback summariser that flags issues before check-out. ### Where the hours come from, per week Front Desk Agents 4 × 1.5h daily AI 6h Guest Email Autoresponder automation 5h Weekly Staff Rota Generator automation 4h Daily Revenue & Occupancy Snapshot automation 3h Housekeepers 5 × 0.5h daily AI 2.5h General Manager 1 × 2h daily AI 2h Front Desk Manager 1 × 2h daily AI 2h Everyone else (7 more) daily AI and smaller automations 12h ### Paid back in week 7 At $32 an hour, the time saved earns back the $7,700 in 7 weeks. Every week after that is profit. Week 1 Week 13 Week 26 1,702h back per year $54,500 value per year Retainer: monthly. The monthly retainer covers maintenance of all automations, updates when Mews or Microsoft changes its tools, monthly one-to-one check-ins with Fiona, and building two new automations each quarter. > Within a week, we turned our biggest time drains into automated processes that feel natural. Now our team spends more time with guests and less time staring at inboxes. The daily report that used to eat my morning before breakfast now arrives while I'm pouring my first coffee. **Fiona MacLeod**, General Manager, The Thistle & Thorn. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "Bayview Insurance Brokers: 31 hours a week back with AI" description: "How a 10-person independent insurance broker for small businesses in Melbourne, Australia got 31 hours a week back in 7 days with Microsoft 365 Copilot: the…" url: https://completeaibusiness.com/case-studies/insurance-broker --- ![A typical morning at Bayview Insurance Brokers: brokers on calls with clients, admin managing schedules, and a whiteboard tracking renewals.](https://completeaibusiness.com/img/cases/insurance-broker-1.webp) Insurance · Melbourne, Australia # Bayview Insurance Brokers got 31 hours a week back in 7 days Bayview Insurance Brokers has helped Melbourne small businesses find the right cover since 2012. We work with over 30 insurers to compare policies and negotiate premiums. Our team of 10 puts client relationships first, cutting through the jargon to deliver clear, actionable advice. A typical morning at Bayview Insurance Brokers: brokers on calls with clients, admin managing schedules, and a whiteboard tracking renewals. **10** people on the team **31h** back every week **$68,400** value per year at $48/h **6 weeks** to earn back the $7,700 ⓘ **We use fictive names for privacy reasons.** Bayview Insurance Brokers and everyone in this story are fictive; the plan shows how a typical 7-day engagement runs in this industry. The business ## Independent insurance advice for Australian small businesses. - **Customers:** Small to medium businesses across Australia: builders, cafes, IT consultancies, manufacturers, and more. - **Team:** 10 people - **Location:** Melbourne, Australia Tools before we started Microsoft 365 (Outlook, Excel, Word, Teams) HubSpot CRM Xero DocuSign Adobe Acrobat BrokerQuote Pro ### What they sell, share of revenue Business Insurance Packages 45% Risk Management Reviews 20% Claims Support & Advocacy 15% Policy Renewal Negotiation 10% Cyber & Professional Indemnity 10% The team ## 10 people, and what their AI now does for them Managing Director AI drafts client proposals and summarises insurer policy changes. 1 Senior Broker AI generates comparison tables from multiple quotes and drafts emails to clients. 2 Account Executive AI handles initial data entry and pre-fills renewal forms. 2 Claims Manager AI summarises claim documents and drafts updates to clients. 1 Risk Consultant AI researches industry risk trends and drafts risk reports. 1 Marketing Coordinator AI writes social media posts, newsletters, and analyses campaign data. 1 Admin/Reception AI sorts emails, schedules appointments, and generates invoices. 2 The questionnaire ## What James Mitchell told us before we started James Mitchell, Managing Director, filled in the same questionnaire you can fill in today. Biggest time drain Manually rekeying client data between our CRM, quoting tools, and policy documents. The goal Give every broker back 10 hours a week so they can spend more time with clients and less on paperwork. Worries about AI AI might give incorrect advice to clients · Staff relying too much on AI and losing product knowledge · Data privacy when using cloud AI tools ### Keep human - Negotiating with insurers on behalf of clients - Walking clients through complex claims - Building long-term relationships ### Take it off our plate - Data entry across multiple systems - Formatting proposals in Word - Chasing missing client information - Sorting and filing email attachments Our advice ## Microsoft 365 Copilot, and the three things to automate first Your team lives in Outlook, Teams, and Excel. Copilot integrates directly into those apps, which means zero new interfaces to learn. It can read your emails, draft replies, build quote comparisons in Excel, and pull client data from your CRM without copying and pasting. Your tools today Microsoft 365 (Outlook, Excel, Word, Teams) HubSpot CRM Xero DocuSign Adobe Acrobat BrokerQuote Pro We advise **Microsoft 365 Copilot** Fit 9/10 Set up for 10 people, connected to your tools AI takes over - Proposal Generator - Email Triage & Filing - Claims Document Summariser - Renewal Data Prep Stays human, AI only supports Final advice and recommendations to clients Complex claims negotiations Building personal rapport with clients ### How each platform fit this business Microsoft 365 Copilot Chosen 9/10 ChatGPT Business 7/10 Claude Team 6/10 Google Gemini for Workspace 5/10 ### What we advised to automate first - Proposal and quote document creation in Word - Email triage and response drafting in Outlook - Data consolidation from BrokerQuote Pro into Excel for comparisons Their 7 days ## One Zoom session a day, from kick-off to AI-ready 1. Day **1** ### Discovery & Audit Managing Director, Senior Broker, Admin Mapped current workflows and identified the highest-friction tasks on Zoom. Agreed on Copilot setup and priority automations. 2. Day **2** ### Platform Setup & Permissions Managing Director, external IT support Configured Microsoft 365 Copilot for all 10 users on Zoom, connected it to their SharePoint and CRM, and set data governance rules. 3. Day **3** ### Individual AI Setup Sessions All staff in rotating 1:1 Zoom calls Held a Zoom call with screen sharing for each team member to configure their Copilot for their role: custom prompts, connected apps, and personalised instructions. 4. Day **4** ### Automation Build Day 1 Senior Brokers, Account Executives Built the first automation on Zoom: a prompt sequence that takes a client's business details and generates a draft proposal in Word with a quote comparison table from Excel data. 5. Day **5** ### Automation Build Day 2 Claims Manager, Admin Automated email filing on Zoom: Copilot now reads incoming emails, classifies them, saves attachments to the correct client folder, and drafts a reply for review. 6. Day **6** ### Training Content Creation Marketing, Admin Recorded custom training videos for each role, covering their most common tasks. Edited and uploaded to their internal knowledge base. 7. Day **7** ### Live Team Workshop & Handover All staff Ran a 90-minute interactive Zoom workshop where the team practiced real scenarios together. Reviewed the automations, gathered feedback, and handed over documentation. ![The Complete AI Business consultant leads a hands-on Copilot session on Zoom with the Bayview team, walking through a real client proposal.](https://completeaibusiness.com/img/cases/insurance-broker-2.webp) The Complete AI Business consultant leads a hands-on Copilot session on Zoom with the Bayview team, walking through a real client proposal. Training ## A live Zoom session, 5 custom videos and a plan per person ### Live training on Zoom We ran a 90-minute hands-on Zoom session with the full team in their meeting room, working through real client scenarios to build confidence using Copilot across Outlook, Teams, and Excel. ### Custom training videos - ▶ Copilot for Brokers: Building a Quote Comparison in Excel 8 min - ▶ Drafting Client Emails with Copilot in Outlook 6 min - ▶ Automating Renewal Letters with Word and SharePoint 7 min - ▶ Using Copilot in Teams to Summarise Client Meetings 5 min - ▶ Admin Essentials: Email Triage and Calendar Scheduling 9 min ### Online training plan, per role - **Managing Director:** Using AI for strategic planning and client communication. - **Senior Broker:** Automating quote comparisons and follow-ups. - **Account Executive:** Streamlining renewal admin and client data management. - **Claims Manager:** Efficient claims document review and client updates. - **Risk Consultant:** Using AI for risk assessment research. - **Marketing Coordinator:** AI for content creation and campaign analysis. - **Admin/Reception:** Automating admin: email, calendar, and invoicing. Automations built ## 4 automations that now run every day ### Proposal Generator **4.5h** per week Before Brokers manually copy-pasted quote data from BrokerQuote Pro into a Word template, typed client details, and formatted tables - taking 45 minutes per proposal. Now Copilot reads a one-line client brief, pulls the latest quote data from an Excel sheet, and populates a branded Word proposal in under 2 minutes. ### Email Triage & Filing **4.5h** per week Before Admin spent 10 hours a week sorting client emails, saving attachments to folders, and flagging urgent messages. Now Copilot categorises incoming emails, saves attachments to SharePoint, drafts a brief reply, and flags priority items in Teams. ### Claims Document Summariser **1h** per week Before Claims Manager read lengthy insurer reports and manually typed summaries for client updates - 3 hours per week. Now Copilot extracts key points from PDF reports and drafts a client-friendly summary email for review. ### Renewal Data Prep **3.5h** per week Before Account Executives re-entered client data into renewal forms and checked policy changes across three systems - 6 hours per week each. Now Copilot pre-fills renewal forms using CRM data and highlights policy changes from insurer emails. The result ## 31 hours a week back, worth $68,400 a year 18 hours come from everyone using their own AI every day, 14 from the automations. The team now handles 20% more client meetings without hiring. Brokers use Copilot daily for drafting, and the admin team has cut email processing time by 75%. ### Where the hours come from, per week Admin/Reception 2 × 2.5h daily AI 4.5h Proposal Generator automation 4.5h Email Triage & Filing automation 4.5h Account Executive 2 × 2h daily AI 4h Senior Broker 2 × 2h daily AI 3.5h Renewal Data Prep automation 3.5h Marketing Coordinator 1 × 2h daily AI 2h Everyone else (4 more) daily AI and smaller automations 4.5h ### Paid back in week 6 At $48 an hour, the time saved earns back the $7,700 in 6 weeks. Every week after that is profit. Week 1 Week 13 Week 26 1,426h back per year $68,400 value per year Retainer: monthly. The monthly retainer covers ongoing Copilot updates, a monthly personal training session for new staff, and building two new automations each quarter. > I was sceptical about AI understanding insurance, but the proposal generator alone saves my brokers hours each week. The real win is that our clients get clearer advice faster, and my team can focus on what they do best - helping people. **James Mitchell**, Managing Director, Bayview Insurance Brokers. Fictive example. ## Your business, AI-ready in 7 days. Answer the questionnaire and get a free first plan, written for your team and your tools. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "Your free AI plan | Complete AI Business" description: "Answer a few questions about your business and team. We research your company and write your first AI plan." url: https://completeaibusiness.com/questionnaire --- Free and without obligation. We only use your answers to write your plan. --- --- title: "Contact | Complete AI Business" description: "Email us, fill in the free AI plan questionnaire, or write to our Hong Kong office. Every engagement runs fully online over Zoom." url: https://completeaibusiness.com/contact --- Contact # Talk to a human. Email is the fastest way to reach us, and the free AI plan questionnaire is the fastest way to get useful answers about your own business. Every engagement runs fully online over Zoom, wherever your team is. ### Email For questions, quotes and anything else. A person reads it and replies, usually within one working day. [**contact@completeaitraining.com**](mailto:contact@completeaitraining.com) ### Business address Unit B, 11/F, 23 Thomson Road, Wan Chai, Hong Kong SAR China Postal address only — all sessions happen in a Zoom meeting, so there is nothing to travel to. ### Start with the free AI plan Answer a short questionnaire and get a written AI plan for your business in a few minutes, free. [Get your free AI plan](https://completeaibusiness.com/questionnaire) --- --- title: "For developers and AI agents | Complete AI Business" description: "Connect an AI agent to Complete AI Business: an MCP server, an A2A agent, a REST API with OpenAPI, WebMCP in the browser and OAuth for a customer's own plans." url: https://completeaibusiness.com/developers --- For developers and AI agents # Connect your agent to Complete AI Business What a person can do on this site, an agent can do for them: read the offer and the case studies, make a free AI plan, and, once the person allows it, read their own plans. No API key needed for the public parts. ### MCP server Streamable HTTP with JSON responses. Public tools need no token. `https://completeaibusiness.com/mcp` [Server card](https://completeaibusiness.com/.well-known/mcp/server-card.json) ### A2A agent JSON-RPC `message/send`, protocol 0.3. Deterministic answers, no open-ended chat. `https://completeaibusiness.com/a2a` [Agent card](https://completeaibusiness.com/.well-known/agent-card.json) ### REST API The same functions as JSON, described in OpenAPI 3.1. `/api/v1/...` [openapi.json](https://completeaibusiness.com/openapi.json) · [API catalog](https://completeaibusiness.com/.well-known/api-catalog) ### OAuth 2.1 For a customer's own plans. PKCE, dynamic client registration, sign-in with an emailed code. [auth.md](https://completeaibusiness.com/auth.md) · [metadata](https://completeaibusiness.com/.well-known/oauth-authorization-server) ### In the browser Every page registers WebMCP tools for an agent running in the visitor's browser: the offer, the case studies, the questionnaire, and the plan the visitor has open. [webmcp.js](https://completeaibusiness.com/assets/js/webmcp.js) ### Pages as Markdown Send `Accept: text/markdown` to any public page and get Markdown at the same URL. There's also an index for language models. [llms.txt](https://completeaibusiness.com/llms.txt) · [llms-full.txt](https://completeaibusiness.com/llms-full.txt) ## MCP tools Tool What it does Needs ` get_offer` Prices, what's included, the 7 days, how to start, the FAQ nothing ` list_case_studies`, `get_case_study` The 9 case studies (fictive examples) nothing ` create_free_plan` Makes a free AI plan for the person you act for, from the questionnaire's questions. Ready in 1 to 3 minutes and emailed to them. nothing ` get_plan_status` One plan by its id: still being written, or ready with the plan the plan's id ` my_plans`, `get_my_plan` The signed-in customer's own plans OAuth, `plans:read` ` request_start` Asks us to start the initial call ($770) or the 7 days ($7,700). Records the request and emails the customer and our team. Nothing is charged. OAuth, `plans:start` Try it with curl: ``` curl -s https://completeaibusiness.com/mcp \ -H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' \ -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_offer","arguments":{}}}' ``` Or with the MCP Inspector: `npx @modelcontextprotocol/inspector`, transport Streamable HTTP, URL `https://completeaibusiness.com/mcp`. ## A customer's own plans: OAuth 1. Register a client at `POST /oauth/register` (RFC 7591). No key needed. 2. Send the customer to `/oauth/authorize` with PKCE (S256). They type the email address they used for their plan, enter the 6-digit code we email them, and choose Allow or Don't allow. 3. Exchange the code at `POST /oauth/token`. Access tokens last an hour; refresh tokens 30 days and rotate on every use. 4. Send `Authorization: Bearer ...` to `/mcp`, `/a2a` or `/api/v1/me/...`. Without it, you get a 401 that points at the [protected resource metadata](https://completeaibusiness.com/.well-known/oauth-protected-resource). A token reaches only the plans made for that email address. Never ask the customer for the code: they type it on our page. The full guide is [auth.md](https://completeaibusiness.com/auth.md). ## What stays private - A plan belongs to the person who asked for it. It is never listed, indexed or put in llms.txt, and its page has no Markdown version. Reading it takes its id, or the owner's token. - Nothing here reaches anyone's email address or phone number, our team's inboxes, the phone agent, or the work we do for paying customers. - Making plans is limited per address, per email and per day, so please don't loop. ## Discovery documents - [/.well-known/api-catalog](https://completeaibusiness.com/.well-known/api-catalog) (RFC 9727) and [/openapi.json](https://completeaibusiness.com/openapi.json) - [/.well-known/mcp/server-card.json](https://completeaibusiness.com/.well-known/mcp/server-card.json) and [/.well-known/agent-card.json](https://completeaibusiness.com/.well-known/agent-card.json) - [/.well-known/oauth-authorization-server](https://completeaibusiness.com/.well-known/oauth-authorization-server) (RFC 8414) and [/.well-known/oauth-protected-resource](https://completeaibusiness.com/.well-known/oauth-protected-resource) (RFC 9728) - [/.well-known/agent-skills/index.json](https://completeaibusiness.com/.well-known/agent-skills/index.json), with one skill: [SKILL.md](https://completeaibusiness.com/.well-known/agent-skills/complete-ai-business/SKILL.md) - [/.well-known/ai-catalog.json](https://completeaibusiness.com/.well-known/ai-catalog.json) and [robots.txt](https://completeaibusiness.com/robots.txt) (with Content Signals) - [/.well-known/http-message-signatures-directory](https://completeaibusiness.com/.well-known/http-message-signatures-directory): the key our [website reader](https://completeaibusiness.com/bot) signs its requests with Questions or a problem with the API: [contact@completeaitraining.com](mailto:contact@completeaitraining.com). --- --- title: "Our website reader | Complete AI Business" description: "What the Complete AI Business website reader fetches, when and why, how its requests are signed with Web Bot Auth, and how to reach us about it." url: https://completeaibusiness.com/bot --- CompleteAIBusiness/1.0 # Our website reader If our bot visited your site, someone asked us for a free AI plan for a business and gave us that website. We read a few of its pages so the plan fits the business. That's all it does. ## What it fetches, and when - Only when a person asks for a free AI plan (on our questionnaire, by phone, or through an AI agent) and names the website. - The home page, plus at most 3 pages on the same site whose address mentions about, team, services, what we do, careers or jobs. - HTML only, at most 800 KB per page, with a 12-second timeout. It follows up to 5 redirects and never fetches private or local addresses. - It reads the text to write that one plan. It does not crawl further, and we don't use your pages for anything else or train models on them. - Because a person asked for that one site, it works like someone opening the page in a browser and does not read robots.txt. ## How to recognise it The User-Agent is: ``` Mozilla/5.0 (compatible; CompleteAIBusiness/1.0; +https://completeaibusiness.com/bot) ``` Every request is signed with [Web Bot Auth](https://datatracker.ietf.org/wg/webbotauth/about/) (HTTP Message Signatures, RFC 9421) and carries three headers: ``` Signature-Agent: "https://completeaibusiness.com" Signature-Input: sig1=("@authority" "signature-agent");created=...;expires=...;keyid="...";alg="ed25519";nonce="...";tag="web-bot-auth" Signature: sig1=:...: ``` The signature covers your host name and the Signature-Agent header, and is valid for 60 seconds. The Ed25519 public key is published at [/.well-known/http-message-signatures-directory](https://completeaibusiness.com/.well-known/http-message-signatures-directory); its `kid` is the `keyid` in the signature. Anyone can send the same User-Agent, but only we can make a signature that checks out against that key. ## Questions or a problem Email [contact@completeaitraining.com](mailto:contact@completeaitraining.com) with the time of the visit and your domain. If you'd rather we didn't read your site, tell us. --- # Complete AI Business auth.md How an AI agent gets access to a customer's own data at Complete AI Business (https://completeaibusiness.com), and what that access allows. The resource server and the authorization server are both https://completeaibusiness.com. ## Who this is for Agents acting for one person: someone who asked for a free AI plan for their business. A token belongs to that person's email address and reaches only the plans made for it. There are no accounts or passwords, and nothing here costs money. ## What needs no credential - The offer, prices, the 7 days and the FAQ: MCP get_offer, GET https://completeaibusiness.com/api/v1/offer, A2A skill offer - The 9 case studies (fictive examples): MCP list_case_studies and get_case_study, GET https://completeaibusiness.com/api/v1/case-studies - Making a free AI plan for the person you act for: MCP create_free_plan, POST https://completeaibusiness.com/api/v1/plans, A2A skill free-plan. The plan is emailed to the address given. Limits: 3 per address per hour through the agent channels (5 per hour in all), 3 per email address a day, and daily ceilings. - Checking one plan by its id: MCP get_plan_status, GET https://completeaibusiness.com/api/v1/plans/{id}. The 12-character id from the report link is the key, so treat it as private. - All pages as Markdown (Accept: text/markdown), https://completeaibusiness.com/llms.txt and the discovery documents ## What needs a token | What | Where | Scope | |---|---|---| | List the customer's own plans | MCP my_plans, GET https://completeaibusiness.com/api/v1/me/plans, A2A skill my-plans | plans:read | | Read one of them | MCP get_my_plan, GET https://completeaibusiness.com/api/v1/me/plans/{id} | plans:read | | Ask us to start the initial call ($770) or the 7 days ($7,700) | MCP request_start, POST https://completeaibusiness.com/api/v1/me/plans/{id}/start | plans:start | | Who the token belongs to | GET https://completeaibusiness.com/api/v1/me | any | Asking to start records a request and emails the customer and our team, who then send a payment link by hand. Nothing is charged through any of this. ## 1. Discover - Protected resource metadata (RFC 9728): https://completeaibusiness.com/.well-known/oauth-protected-resource, and for the MCP endpoint https://completeaibusiness.com/.well-known/oauth-protected-resource/mcp - Authorization server metadata (RFC 8414): https://completeaibusiness.com/.well-known/oauth-authorization-server (with an agent_auth block) ## 2. Register a client (anonymous, RFC 7591) ```http POST https://completeaibusiness.com/oauth/register Content-Type: application/json {"client_name": "Your agent", "redirect_uris": ["https://your.app/callback"], "token_endpoint_auth_method": "none"} ``` The reply has your client_id (and a client_secret if you asked for client_secret_post or client_secret_basic). Redirect URIs can be https, http on a loopback address (127.0.0.1, localhost, [::1], any port), or a native app's private-use scheme. 20 registrations per address per hour. ## 3. The person signs in and approves (the claim) Open this for the person, in their browser: ``` https://completeaibusiness.com/oauth/authorize?response_type=code&client_id=CLIENT_ID&redirect_uri=REDIRECT_URI&scope=plans:read&state=STATE&code_challenge=CHALLENGE&code_challenge_method=S256 ``` - PKCE with S256 is required. Add plans:start to the scope only if the person may ask us to start. Optional: resource=https://completeaibusiness.com/mcp (RFC 8707); such a token then works only there. - The person types the email address they used for their plan. We email them a 6-digit code (10 minutes, 5 tries). Then they see what your app asks for and choose Allow or Don't allow. - Never ask the person for the code. They type it on our page, not in your app. - Codes are limited to 10 per address and 5 per email an hour. You get the person back at your redirect_uri with code, state and iss=https://completeaibusiness.com. ## 4. Exchange the code ```http POST https://completeaibusiness.com/oauth/token Content-Type: application/x-www-form-urlencoded grant_type=authorization_code&code=CODE&redirect_uri=REDIRECT_URI&client_id=CLIENT_ID&code_verifier=VERIFIER ``` The reply: access_token (1 hour), refresh_token (30 days) and scope. Refresh with grant_type=refresh_token. Each refresh token works once and is replaced; using an old one revokes every token from that sign-in. Codes work once, for 10 minutes. We store only hashes of codes and tokens. ## 5. Use the token Send Authorization: Bearer ACCESS_TOKEN to https://completeaibusiness.com/mcp, https://completeaibusiness.com/a2a or https://completeaibusiness.com/api/v1/me/... Without a valid token you get 401 with WWW-Authenticate: Bearer resource_metadata="...". A token without the needed scope gets 403 with error="insufficient_scope". A token that is sent must be valid, even for public tools. ## 6. Revoke ```http POST https://completeaibusiness.com/oauth/revoke Content-Type: application/x-www-form-urlencoded token=TOKEN&client_id=CLIENT_ID ``` Revoking a refresh token ends the whole sign-in. The person can also email contact@completeaitraining.com to end an app's access. ## What no token can reach Other people's plans, any list of plans, email addresses or phone numbers of others, our team's inboxes, the phone agent's endpoints and memory, how plans are written, and anything we make for a paying customer.