Let’s talk

Case study · Accounting and advisory firm · Midwest US

Client meeting notes turned into a weekly brief for a fractional CFO team

AI agents read the firm’s client meeting notes and give each CFO a short weekly brief: overdue actions, the conversations to have next, frustrations building and signals to watch, with a bot to ask before a call.

3 to 5 minweekly read per client, in place of months of meeting notes
Business
Accounting and advisory firm · Midwest US
The problem
The insight was in the meeting notes. Nobody had time to read them.
The result
The first agents are live and in daily use by the CFO team.
Services
Operational Improvement Project, Knowledge Capture and Handover

The situation

The insight was in the meeting notes. Nobody had time to read them.

A regional accounting and advisory firm provides fractional CFO services to owner-led businesses. Each CFO looks after a portfolio of clients, and every client meeting produces notes: actions agreed, worries raised, decisions put off.

Those notes were written and rarely read again. Action items slipped, a client's repeated frustration went unnoticed until it became a problem, and preparing for a call meant scrolling back through months of notes. The firm's leaders wanted the same approach for coaching their own people, and they did not want a large one-off project fee every time they had a new idea.

What we did

Five steps, in order.

  1. Step 01

    Started with one data source, the client meeting notes, and one audience, the CFOs, rather than connecting email, calendars and chat all at once.

  2. Step 02

    Agreed what each CFO would read in a three to five minute weekly brief: outstanding actions grouped by urgency, the three conversations the client needs next, frustration and tension points shown as patterns rather than one-off remarks, good questions to ask, and early-warning signals such as cash stress or decision avoidance.

  3. Step 03

    Built the first agents and switched them on, with each CFO seeing only their own clients.

  4. Step 04

    Scoped the next three with the firm: a client bot inside Microsoft Teams that answers questions such as "What am I avoiding with this client?" and "What should I prepare before tomorrow's call?"; a training bot that turns a video or document into a reflective conversation ending in a personal action plan; and a coaching bot that turns review feedback into a weekly doing-well, slipping, ask-next report for each coach.

  5. Step 05

    Set the arrangement up so the firm's own team can build and adapt agents on the platform over time, instead of paying a project fee for every new idea.

What changed

The first agents are live and in daily use by the CFO team.

  • Each CFO gets the actions, tensions and questions for a client in a short weekly read, instead of a pile of notes.
  • Three further agents scoped, with the firm's own people able to extend them.

What the team uses now

The CFO team reads the weekly brief and the firm is moving on to the Teams client bot and the coaching agents. The agents run on Decoder, the AI platform founder Peter Dolukhanov also founded.

The services behind it

The same steps, available for your business.

Operational Improvement ProjectDeliver an agreed improvement, bring the team along and measure the results.Typically 6 to 12 weeks, in phasesSee what’s includedKnowledge Capture and HandoverCapture experienced people's know-how, document the procedures that matter, and make it easy to find and use.4 to 6 weeksSee what’s included
Talk to us about a similar problemAll results

Free calculator

What would a result like this be worth in your business?

Put in how many people do document and admin work and how long it takes them. See the cost a year and what the staff time a first automation could free up is worth. Nothing is stored.

Try the calculator

More case studies

Independent insurance agency · Southern CaliforniaDocument and admin work automated for the agency's staff

AI now reads incoming documents, pulls out the details and prepares proposal paperwork across the agency's systems, carrier portals and email, so staff stop re-keying the same information.

Read the case studyTesting laboratory · CaliforniaA laboratory system a year late, back on track

We took over running the rollout with the vendor, renegotiated the agreement so payments follow delivery, and designed the tests the lab's own team used to check the system before accepting it.

Read the case studyManufacturer · Southern CaliforniaOne consistent way to quote jobs at a family-owned manufacturer

As on-demand CTO for a new operations system, we designed how estimates and pricing work and used AI to convert years of old estimates, so the team quotes from the same data.

Read the case studyCreative industry association · New York and EuropeA conference captured, analysed and reported by AI in real time

Live transcription of every keynote, breakout and interview, an AI agent pulling quotes and themes onto the venue screens as they happen, and a branded post-event report with takeaways within a week instead of a written summary weeks later.

Read the case studyCustom fabricator · Southern CaliforniaThe right business platform for a fabrication shop, chosen independently

Requirements written down with the owner, 13 platforms scored against them, and an independent recommendation that fits a shop rather than a general contractor, so the team can work without everything flowing through the owner.

Read the case study

Let’s talk

Tell us about the problem. You'll hear back within one business day.

Let’s talk