Natural language data querying is no longer a future promise — it’s an engineering problem you can solve today. In this talk, we’ll walk through how to combine Apache Gravitino’s unified metadata management with dbt’s MetricFlow to construct a robust semantic layer that sits between your raw data and your AI agents. You’ll learn how Gravitino catalogs and governs your data assets across heterogeneous sources, how MetricFlow defines consistent, reusable metrics on top of them, and how to wire both together so that a data agent can answer business questions reliably — without hallucinating schema details or misinterpreting metric definitions. We’ll close with a live demo of an agent asking questions against a multi-source environment and getting accurate, governed answers.

Speakers:

Jerry Shao: Datastrato, CTO

Jerry Shao is the co-founder and CTO of Datastrato, focused on open source Big Data are for more than 10 years. He is an Apache member, committer and PMC member of Apache Spark and Apache Inlong, the original creator of Apache Gravitino.