Towards Agentic Schema Refinement
This addresses the challenge of data exploration for users in enterprise settings, though it appears incremental as it builds on existing LLM and database techniques.
The paper tackles the problem of complex enterprise databases obscuring data semantics by proposing a semantic layer of interpretable database views, using a multi-agent LLM simulation to discover these views with minimal input.
Large enterprise databases can be complex and messy, obscuring the data semantics needed for analytical tasks. We propose a semantic layer in-between the database and the user as a set of small and easy-to-interpret database views, effectively acting as a refined version of the schema. To discover these views, we introduce a multi-agent Large Language Model (LLM) simulation where LLM agents collaborate to iteratively define and refine views with minimal input. Our approach paves the way for LLM-powered exploration of unwieldy databases.