Agentic AutoML on Relational Databases

Most structured data lives in relational databases, but AutoML still assumes a single flat table — so humans spend their effort flattening. This project surveys AutoML and relational deep learning, sketches a formalism for agents that explore schemas, enrich features and search models directly on the database, and asks whether copy-on-write database clones are the right substrate for the many experiment versions such agents create.

Ouael Ben Amara

Project lead · university of Michigan