This is a short recording from one of our internal working sessions. Kelvin walks through the workforce-first framework we use on every Responsible AI engagement — four outcomes, one example, five minutes.
Transcript · 00:00
"Most enterprise AI fails because the engagement starts with the wrong question. The wrong question is which roles can I cut. The right question is which work, today, is making my best people slower than they should be — and how do I give them the AI that fixes that…"
What's in the video.
- 00:00 – 01:10 · Why the headcount-first framing fails.
- 01:10 – 02:35 · The four outcomes we measure instead.
- 02:35 – 04:00 · One example from a Responsible AI engagement in financial services.
- 04:00 – 05:12 · How we phrase this on the first call with a CIO.
If you only watch one minute.
Skip to 02:35. That's the example. Most people understand the framework instantly once they see it applied to a real workflow.
Note
This is a working session, not a polished talk. The audio is decent. The handwriting is mediocre. The framework is good.
— Kelvin Tran. Los Angeles, May 2026.
Comments