Before the agents: what self-serve analytics actually needs

Self-serve analytics tools built with AI look identical from the outside — the agents, the eval harness, the querying workflow all copy easily. What sits underneath does not. In this session, KOHO walks through how sustainable self-serve rests on a business context layer built from real questions and the analytics engineering practice in dbt that makes a metric definition trustworthy in the first place — plus measured results comparing analyst-authored context against AI-generated context on an identical benchmark.