# Data + AI Summit 2026, Through a Governance Lens > Source: > Published: 2026-07-08 00:01:02+00:00 Member-only story # Data + AI Summit 2026, Through a Governance Lens I spend most of my working life thinking about lineage, access policies, and audit trails, and about the dozen quiet ways a “quick pipeline fix” can wreck a governance model six months down the line. So when I sat down to go through Databricks’ Data + AI Summit 2026 (June 15–18, Moscone Center), the individual product names mattered less to me than the argument sitting underneath all of them: your agents can already reach your data. They still don’t understand your business. Every headline launch this year, once you squint at it, was Databricks trying to close that exact gap, just from a different angle of the stack. Here’s my take on the four announcements that matter most, plus a smaller side-by-side with what Snowflake shipped at its own summit two weeks earlier. ## LTAP: the infrastructure swing LTAP, short for Lake Transactional/Analytical Processing, is the big infrastructure bet this year. It’s built on Lakebase, Databricks’ Postgres service, and the pitch is that operational and analytical data live in the same open formats, Delta and Iceberg, inside Unity Catalog. No CDC pipeline shuttling data between Lakebase and the Lakehouse. Nothing to maintain. Nothing quietly drifting out of sync while nobody’s watching.