What if I told you you’re saying goodbye to 70 full working days of important data without realizing it?
The average manager sits in 564 meetings per year.
More than a quarter of the year spent in conversations. And for a CFO? It is probably closer to 25 or 30 hours every single week between investor calls, board sessions, business unit reviews, and budget negotiations.
Meetings where someone flags a risk, someone says something smart about changing headcount, your CEO mentions their changing priorities.
All of these conversions are full of useful intelligence for your financial planning.
But most of it never makes it into your model, your forecast, or your risk register.
Because, It disappears the moment the meeting ends.
Memory is Not A Reliable System
|
|
I see this everywhere when I train finance teams. Somebody takes notes during a meeting. Maybe the notes are in a notebook, maybe in a random OneNote page, maybe in an email to themselves. And by next week, nobody can find them or remember the context.
But, none of this shows up in your ERP, and none of it appears in your variance analysis.
It only exists in the memory of whoever was in the room.
And memory is not a reliable system.
When I talk to the best finance leaders, the ones who seem to know about problems before they hit the numbers, they all have one thing in common. They treat meetings as a data source. Not just a communication event where people talk and then go back to their desks. A data source, like your ERP or your CRM, that needs to be captured and used.
After my discussion with AI CFO – Saul Mateos – my team wrote up this in one of our AI Finance Club articles:
"No meeting catches him off-guard, and he can walk into each discussion armed with key facts and context that would have taken considerable time to compile manually. The AI essentially created a new layer of efficiency and intelligence in his morning workflow."
You can learn more about his exact morning routine in my previous newsletter here.
Building a Meeting Intelligence System
|
|
So, the question is, how do you go from losing all of this information to having it work for you?
The answer is something I have been building with several finance teams, and it is simpler than you probably think.
You connect your meetings to a central knowledge base, and then you use AI to extract the things that matter from every conversation.
Once this system is running, you can ask questions that are currently impossible to answer. Things like:
Or,
Those are the kinds of questions that separate a finance leader who is reacting to surprises from one who sees them coming three months early.
4 Steps to Get This Running
|
|
Step 1 — Turn on automatic transcription
Connect an AI notetaker to every meeting. If you are on Microsoft 365, Teams already has built-in transcription. It’s the same with Google Meet.
Otherwise, tools like Otter.ai, Fireflies, or Zoom AI all work.
The goal is simple: every meeting produces a transcript, saved automatically to a shared folder. No manual work required from anyone.
Step 2 — Build your extraction agent
Create a Claude Project or Custom GPT and connect it to your transcript folder (make sure you are using a secure AI tool).
|
|
Load it with the context it needs to understand your business: your key metrics, the names of your business units, your current strategic priorities, and your existing risk register.
This is what allows the AI to recognize when something in a transcript is relevant, not just factually, but strategically.
Step 3 — Define what you want to extract
Every transcript should run through the same four filters.
- Decisions made: what was formally agreed?
- Risks flagged: what concerns were raised, even informally?
- Commitments given: who promised what, and by when?
- Strategic signals: what was said that might affect the financial plan, even if it was not framed as a financial issue?
Step 4 — Run a weekly synthesis
Every Friday morning, run one prompt across the full week of transcripts:
This takes about 10 minutes. And it surfaces things that would have required you to remember every conversation from the past five days and connect the dots manually in your head.
Here's a sample of AI summarizing the latest transcript from a Google Drive folder.
|
|
Want to Build This Faster?
Here's a free step-by-step guide that you can download instantly from this e-mail as a PDF.
📥Click here 👉 Financial Review Extractor Agent Handbook.pdf
|
|
The One Thing to Remember
Meeting transcripts are data.
Conversations are constantly happening, and insights are constantly emerging.
And better visibility of this will change how you advise your CEO, how you present to the board, and how confident you feel walking into your next planning cycle.
70 days of meetings… The intelligence was always there, now it's yours to use strategically.
Best,
Your AI Finance Expert,
Nicolas
P.S. — What did you think of this one? Hit reply and let me know if you are going to try this (I read all replies).
P.P.S. — If you want to see the meeting prep system I showed recently, check out my latest YouTube video, "How This CFO Uses AI to Save 10+ Hours Per Week" here.
link