Question

A Generative AI Engineer maintains a production RAG application whose Vector Search index is built from a Delta table, and whose serving chain is a registered model in Unity Catalog. The data science team plans to replace the embedding model used to build the index, and leadership wants to know exactly which downstream assets (indexes, registered models, and serving endpoints) would be affected before any change is made. What is the most efficient way to determine the full set of impacted downstream assets?

Accepted Answer

Use Unity Catalog lineage (via Catalog Explorer or the lineage system tables/API) to trace the downstream dependencies from the source table and embedding assets to indexes, models, and endpoints.

Answer

Manually inspect each notebook and job in the workspace to find references to the embedding model and record the assets they touch.

Answer

Enable inference tables on the serving endpoint and analyze logged payloads to infer which assets depend on the embedding model.

Answer

Query the MLflow experiment run history to find all runs that logged the embedding model as a parameter.