For teams and for events
The team has the AI licences. Ask what has gone into the board pack since, and the answer is nothing. The output is there. Nobody is willing to sign it.
What they leave with
The same method as the 1 on 1 coaching, run for the whole room. 4 steps, on the team's own work.
Every recurring decision the team owns, on one page: who decides, and on what evidence.
Each decision gets one of the 3 marks. The argument that never happens otherwise happens here, and the trust comes from it.
What the workflow may never decide alone, what has to be traceable, and which cash limit sits above the model.
The team builds one workflow in the room and runs it against numbers it already knows are right.
Who teaches it
I have sat in the CFO chair, and at 3M I ran inventory operations that cut non-moving inventory from 8.5% to 5% while holding a 100% fill rate through the semiconductor shortage. I coach inside the AI Finance Club Accelerator: live sessions and written reviews of the projects finance leaders bring.
I've seen how she breaks down technical AI and data concepts into something finance professionals can actually apply. Her blend of operational finance background and AI fluency is rare, and it makes her advice both credible and immediately useful.
Charles T.CEO, intellectual property technologyShe doesn't critique outputs, she diagnoses the system behind them: backtesting, train/test splits, auditability. That's the rigor senior finance professionals actually need.
Albert LeeFounder, FP&A advisory practice. FCPAShe has great ability in distilling complex subject into practical workflow and application.
Emily MooreSenior finance and FP&A leader. CPA, PMPShe offered insightful and timely responses to student queries during live sessions, and offered 1:1 support to all students within the cohort. She is knowledgeable, supportive and humble: a rare combination.
Three formats
Open the one that fits. Each is scoped to the event, and the proposal comes back on one page.
Speaking: keynote or talk
The talk asked for most: Designing AI workflows that improve decisions, not just speed. AI is producing more output, and decisions are not getting better. Three gaps sit behind that: no decision logic, no financial boundaries, and no record of why a person overrode the machine. It lands through scenes the room has lived: the reorder finance rejected because cash was never in the model, the forecast nobody trusted, the recommendation quietly ignored.
Other talks
Workshop: a working session for one team
Laptops open, on the team's own numbers. Finance, procurement and operations in the same room works best: the argument about whose number is right comes out and gets settled.
Group coaching: one finance team, over several weeks
The same coached method as the 1 on 1 build, run for a whole team. Each person builds on the team's own data, and each workflow is tested against numbers the team already knows are right before anyone relies on it. Scoped to the size of the team and what rides on the numbers.
Virtual, worldwide. In English or Mandarin.
Tell me who is in the room, how many, and what they should do differently afterwards. You get a one-page proposal: the format, what it covers, and the fee.