
There are two unclimbable mountains in database world: 1 Combined transactional/analytical processing 2 true multi-master scale out We are making significant progress on both in Lakebase

VP @databricks, Founder @neondatabase. Building Lakebase Postgres for AI-native world

There are two unclimbable mountains in database world: 1 Combined transactional/analytical processing 2 true multi-master scale out We are making significant progress on both in Lakebase

One click deploy to @vercel, @neondatabase and @resend - signudoc.app
DocuSign wants $8,100/yr from a 15-person team just to sign PDFs. So I built SignuDoc: open source, one-click deploy to your own @vercel account, $0 per seat. White-label branding, full audit trails, tamper-evident seals and your documents never leave your cloud.

add-mcp surpassed 400k weekly downloads! 🎉

You can build agents with a full backend on @neondatabase and still own every piece of your stack. All our primitives are built on standard interfaces and fully composable. We don’t force infra you don’t want
The @ideavo_ai agent creates a Neon project for every app through the API, and users can claim that database into their own Neon account whenever they’re ready

Another packed Neon changelog today: - Postgres minor version update - Claude Fable 5.2 and Kimi K3.1 Pro on AI Gateway - Stripe to Postgres sync integration - Project permissions for Neon backend services - New Functions templates - Updated agent skills neon.com/docs/changelog/2026…

Welcome back from SPACE!
You're not gonna believe this, but @SpaceX caught their Starship rocket at 4:20 PM ET 😂

Changelog on tap
Another packed Neon changelog today: - Postgres minor version update - Claude Fable 5.2 and Kimi K3.1 Pro on AI Gateway - Stripe to Postgres sync integration - Project permissions for Neon backend services - New Functions templates - Updated agent skills neon.com/docs/changelog/2026…

Tonight the Postgres security team released minor version updates patching 52 CVEs in Postgres. Neon and Databricks customers were running patched versions within 30 minutes thanks to Lakebase architecture and our Postgres engineering team. Always-on computes will get the updates on next restart. neon.com/docs/postgresql/pos…

Operational excellence from the Lakebase team. Both @neondatabase and Databricks customers running latest minor version within 30 minutes of release.
Tonight the Postgres security team released minor version updates patching 52 CVEs in Postgres. Neon and Databricks customers were running patched versions within 30 minutes thanks to Lakebase architecture and our Postgres engineering team. Always-on computes will get the updates on next restart. neon.com/docs/postgresql/pos…

Lakebase will mark a phase shift in database technology! It'll enable LTAP, true multi-master, etc.
There are two unclimbable mountains in database world: 1 Combined transactional/analytical processing 2 true multi-master scale out We are making significant progress on both in Lakebase

Transactional/analytical is being solved with LTAP and we will swing at multi-master super hard. I'm excited to be able to solve such long standing database problems.

It all changes with agents and we feel we are in a unique point in time where we can be a lot more ambitious in taking on giant systems project with agentic software production.

In addition to that every experience database kernel engineer knows that multi-master implementation will take years and only then you'll start learning if your service is any good.

I believe that multi-master approach pioneered by Oracle RAC should be revisited for the cloud era. Nobody has been able to build a successful multi-master cloud service so far. Aurora cancelled their due to problems with optimistic concurrency and no other multi master implementation became popular.

I already spoke about scale out here: x.com/nikitabase/status/1723… and I'm not a fan of shared nothing architectures. But if not shared nothing, what SHOULD be the scale out approach
Database architecture thread. Technical. There has been several startups building an operational relational databases focused on OLTP with a shared nothing architecture. @neondatabase is using a different approach - shared storage. What's the difference?

On (1) we call it LTAP. The idea is that we can leverage Lakebase storage architecture to have the slowest tier represented in Iceberg data format. docs.databricks.com/aws/en/o…