TLDR Tech / Practical
Multi-Tenant Data Agents: Serving Many Customers From One Lakehouse
Building a multi-tenant data agent on a single lakehouse requires isolating data, semantics, and agent state across multiple components rather than relying solely on query filters or prompt instructions. The author outlines three data isolation models, mapping tenant identities to Apache Polaris catalogs and access controls, and carrying authenticated tenant context through session-bound tools to prevent cross-tenant data leaks.
Read the full story