Deploying AI in a bank isn't an engineering problem — Paden Gayle on approvals and evidence packs
AI Engineer · youtube · 2026-10-13
Applied AI engineer Paden Gayle uses a hypothetical bank document-exception assistant to show why a working demo still can't ship: the missing decisions are about where the system runs, who authorizes documents, who reviews wrong exceptions, and which controls apply. His method: find decision owners in week one; shadow, draft and shrink the workflow with control owners; grant the least authority first while the assistant gathers evidence on errors and overrides; and assemble a five-file evidence pack (intended use, data lineage, failure modes, per-segment results, accountable controls), since aggregate metrics can hide weak segments and reruns can't reconstruct history.
More from coding & agent
- Chrome team untangled 6M lines of Chromium legacy code in weeks with Gemini 4 Argon — tulseedoshi · 2026-10-14
- Chrome team used Gemini 4 Argon to untangle 6M lines of Chromium code in weeks — tulseedoshi · 2026-10-14
- Sazabi 1.0 launches: an observability safety harness for teams shipping with agents — henloitsjoyce · 2026-10-14
- Phil Schmid: Gemini 4 Argon drove a months-long Chrome refactor at ~4x usual pace — _philschmid · 2026-10-14
- Coding agent realizes it's in the wrong project, finds the right one on its own, and ships the fix — dfinke · 2026-10-14
- Coding agent announces it 'located' the component it was already inside — dfinke · 2026-10-14