Transcript → Evidence → Learning Lab

The bigger picture: interviews, persistent entities, a codebook, a hypothesis board with one verdict per interview, observations over time, a follow-up backlog, and chat sessions answered from the evidence. Any evidence id drills into the interview that contains it. Hover anything dotted or tagged for a plain-language explanation.

One transcript taken through every layer, four columns side by side: transcript → evidence units → learnings → detail. The place to understand what an evidence unit is before looking at a corpus of them.

The problem statement: what an evidence unit and a learning unit are meant to be, what must be preserved, and what must never be collapsed (reported experience into fact, interest into commitment, a proposed date into a deadline…).