With Eugene Wu of Columbia's DAP Lab, on what it takes to trust AI agents in regulated environments. Why building an agent is easier than validating it.

The four barriers that stall enterprise AI after the pilot. Purple Fabric's answer, on record.

A live multi-agent demo took a complaints case from five weeks to twenty minutes in front of 500 bankers. Attendees said it changed how they judge vendors.

Spoke on fully compliant AI for consumer duty at the FT Adviser conference in London. The only speaker given a dedicated video highlight.