Databricks launches Lakebase Search with full-text and vector search natively in Postgres for AI agents
postgres
39 tracked signals on postgres.
Electric joins Databricks to bring WASM Postgres to AI agent sandboxes
Databricks built Lakebase, a serverless Postgres that decouples compute from storage by moving storage to the lake.
“we're going to have more software generated in the next year than in the history of humanity”
Databricks unveils LTAP, combining Lake Base (managed Postgres) and a transactional/analytical engine for real-time agentic data workflows.
“So Lake Base is open source Postgres fully managed by Databricks with separated storage and compute and it's really easy to get started.”
Databricks introduces LTAP, a storage technology unifying OLTP and OLAP on one copy of data.
“We think Altab is H-Tab done right and accomplished the goals of H-Tab without actually having a single query engine for it, but we able to actually unify the storage which is by far the most important part.”
Databricks Lakebase reimagines Postgres with object storage and WAL for agentic AI workloads
“Agents that interact with a traditional OLTP database often create bottlenecks at the storage layer.”
AI agents move data retrieval from single-turn RAG to multi-step tool use, but agent-generated SQL often fails retrieval.
“why 90% accuracy in text to SQL is 100% useless”
AI-based Postgres tuning needs built-in safety brakes to prevent dangerous configurations from destabilizing production systems.
“Fast Cars Need Fast Brakes.”
Databricks positions Lakebase Postgres and agentic AI as key drivers for vertical industry transformation
Databricks Lakebase Postgres simplifies AI agent orchestration for enterprise workflows
pg_lake, an open-source extension turning Postgres into a lakehouse, now sits at the core of Snowflake's Postgres service.
“pg_lake essentially brings this technology into Postgres.”
Microsoft launched HorizonDB public preview and made 345 Postgres 19 commits while advancing AI-enhanced database tooling.
“In Postgres 19, Microsoft made about 345 commits to the upstream project and changed around 64,000 lines of code.”
Microsoft is expanding its Postgres developer tooling beyond VS Code to support Cursor IDE and emerging AI agent workflows.
“development isn't just happening in editors anymore. The work is moving into new terminals with AI-driven tools, and our goal remains the same. We want to meet developers where they are.”
Microsoft announced Azure HorizonDB, a new Postgres service in public preview at Build.
“what we're trying to do with Horizon DB is really serve a very large market by bringing a bunch of new AI capabilities and developercentric tooling in VS Code”
Databricks Lakebase now autoscales Postgres instances to match workload demand.
Supabase became a decacorn within a few years of founding as a Postgres hosting tool.
“we should not exist if AWS are so good. They've got way more cash than us”
New Citus version adds parallel rebalancing to speed up cluster scaling when adding worker nodes.
“The data was not a bottleneck. The machinery that moves the data was the bottleneck.”
CloudNativePG's new failoverQuorum feature brings quorum-based consistency to synchronous Postgres replication on Kubernetes.
“if you're running Postgres on Kubernetes, then I do think everybody should use synchronous replication”
Local NVMe storage runs Postgres ~100x faster than EBS by eliminating the disk I/O bottleneck behind most performance problems.
“Postgres isn't slow. Your storage is.”
Apache AGE, a Postgres extension, brings graph workloads and visualization on top of relational data stores.
“A graph is just simply a way of thinking about data that prioritizes connections. A graph database is a system built to store and query that structure at scale.”
Microsoft trains 200+ field engineers on Postgres using modular, persona-mapped content rather than one-size-fits-all curriculum.
“Postgres is the database that never changes phone number.”
Microsoft showcases its Postgres contributions and Postgres AI tooling like MCP servers in the POSETTE 2026 event trailer.
“We're going to look at the range of ways you can build MCP servers from very exploratory servers where the agents can send full SQL queries to your database”
Databricks Lakebase introduces branch-based Postgres restores to speed up OLTP recovery at scale
Databricks Lakebase offers cost optimization strategies for managed Postgres workloads
Discrete event simulation can model Postgres CPU-credit degradation on burstable cloud instances to shift-left performance testing before deployment.
“So actually you will go into the death spiral where it will keep on burning your excess CPU and you can't recover.”
A Cohesity engineer argues Postgres security must be a proactive design habit, not a reactive afterthought.
“Is your Postgres really secure or is it just working?”
POSETTE 2026 Livestream 3 previews Postgres talks spanning WAL internals, logical replication, performance, and AI agents on databases.
“The move from simple queries to AI agents.”
Postgres 19 enters feature freeze with self-tuning async I/O worker pools as a headline improvement.
“Andres Freund reviewed my patch and he looked at it and said, it seems to work really well, but I have no idea why.”
Mapping Postgres shared buffers with huge pages drastically cuts page-table memory overhead when scaling to many connections.
“we shall stop guessing and start tuning our systems”
Microsoft built Azure DocumentDB by open-sourcing its managed MongoDB-compatible, Postgres-backed engine and feeding learnings back into the service.
“we started with a managed service, went back into building an open source engine from the managed service, but then funneled all of our open-source learnings from the project back into the managed service”
A practical guide arguing range-based partitioning is the only good way to partition large Postgres tables.
“If your tables are around 100 gigabytes, that's a good moment to start to think about partitioning.”
Postgres NOTIFY in triggers forces accidental serialization via a global pg_database lock, stalling concurrent commits.
“Postgres treats LISTEN/NOTIFY like a globally serialized queue of notifications”
A POSETTE 2026 talk explains how a Postgres INSERT flows through WAL, logical decoding, and replication slots to a subscriber.
“We start from the insert, the time the insert begins on the publisher, makes its way to the shared buffer, eventually to the WAL.”
POSETTE 2026 Livestream 4 previews talks on Postgres partitioning, SQL/PGQ graph queries, logical replication, and PG DuckDB analytics.
“SQL PGQ is the latest addition to SQL standard.”
Postgres' write-ahead log underpins crash recovery, point-in-time recovery, and physical and logical replication.
“Write-ahead log is basically the change-log for the database.”
Postgres uses pg_statistic stats like correlation and most-common-values to decide between sequential and index scans.
“Now you might be tempted to think that something's broken, but it's actually working just fine.”
Postgres 18 introduces temporal keys and promotes NOT NULL to a first-class constraint.
“I read all the changelog, so you don't need to go through it yourself.”
POSETTE 2026 Livestream 2 previews Postgres talks on the feature freeze, PG Lake, Oracle migration, internals, and authentication.
“Why do some Postgres patches make the cut while others miss the boat or even get reverted in the end?”
A Postgres talk examines whether random_page_cost's 25-year-old default of 4.0 should be lowered for modern SSD storage.
“Sequential I/O is much faster, like, than random I/O. And that is true even on SSDs.”