events

Agentic Memory on Databricks: Deploy DB Agent with Databricks Lakebase

The Toronto AI Meetup hosts an online session on deploying the open-source DB Agent on Databricks, with Lakebase acting as both the operational database and, through its native pgvector extension, the vector store behind agentic memory — on

In inglese

Starts

2026-10-12T21:00:00+00:00

Organizer

Toronto AI Meetup

Location

Online event

The Toronto AI Meetup hosts an online session on deploying the open-source DB Agent on Databricks, with Lakebase acting as both the operational database and, through its native pgvector extension, the vector store behind agentic memory — one instance rather than a separate vector service. The talk centres on a redaction-first pattern: every interaction passes through a second LLM call that writes a sanitized summary, entities by ID and type rather than literal values, so agents in different security zones share context without exposing the underlying data. A silent embedding bug found during testing and its fix are also covered.