Just-in-Time Ontologies: What Cheap Intelligence Means for Data — Hrishi Olickel, Southbridge AI

Why watch

Hrishi Olickel explains why universal schemas and confident agents fail at data transformation—using a $784K→$5.3M debt bug as motivation. He introduces just-in-time ontologies: build transformations when the question is known, preserving context and evidence. The Ohio campaign finance entity resolution project demonstrates separating pipeline-design models from execution models, using disagreement checking between model families, and establishing stable execution surfaces with bounded budgets. For data engineers and ML engineers building production data pipelines.