In 30 seconds
Supabase’s founders discuss Compute, Postgres infrastructure, the Turso acquisition, live migrations and declarative schemas. Host swyx discloses a seed investment and podcast sponsorship by the company.
- 1ComputeSupabase describes long-running compute near its database.
- 2Postgres infrastructureThe company discusses OrioleDB and Multigres testing and releases.
- 3TursoThe Turso acquisition supports its agent database plans.
- 4Early launchA positioning change brought an early surge of users.
- 5AI adoptionVector support and AI builders contributed to adoption, the founders say.
- 6Live migrationLive migration must manage writes and consistency.
- 7Declarative schemasA declarative schema gives an agent a direct representation of desired structure.
Takeaways
- Supabase describes long-running compute near its database.
- The company discusses OrioleDB and Multigres testing and releases.
- The Turso acquisition supports its agent database plans.
- A positioning change brought an early surge of users.
- Vector support and AI builders contributed to adoption, the founders say.
- Live migration must manage writes and consistency.
- A declarative schema gives an agent a direct representation of desired structure.
Read the edited text by chapter
Compute and database infrastructure
Product announcements and scale.
Latent Space host swyx interviews Supabase co-founders Paul Copplestone and Ant Wilson after the company’s 2026 Select announcements. He discloses that he invested in Supabase’s seed round and that the company sponsors the podcast, so the conversation has a financial relationship.
The founders present long-running Supabase Compute alongside the database. They describe running container workloads on the same private network as Postgres, extending beyond the earlier edge-function use case. These are product descriptions from the company.
They discuss OrioleDB, testing and a beta release under the Apache 2 license, as well as Multigres for managing and scaling Postgres. Their reliability claims reflect their own tests and should not be read as a guarantee that data loss is impossible.
The founders explain the Turso acquisition as part of preparing for large numbers of agent-driven databases. They contrast scaling a single database with distributing many databases and discuss SQLite’s single-writer model.
The conversation shifts to developer marketing and community. The founders describe using humor and public communication while building a distributed company.
Launch and growth
Positioning, vectors and AI builders.
They recount an early positioning change from “real-time Postgres” to an open-source Firebase alternative, followed by a sudden influx of users. That launch exposed infrastructure and migration problems that the team had to solve.
The founders describe vector support in Postgres as an early AI-related growth driver, followed by demand from AI application builders. They argue that recommendations by agents can influence which backend developers choose; the extent of that effect is a company and host interpretation.
Enterprise and migration
Sales signals and consistency.
On enterprise growth, they describe product-led sales based on signals such as usage, limits and errors. They discuss comparing contacted users with a nonresponding group; that comparison alone does not establish a randomized causal effect.
The conversation turns to live database migrations. Moving a transaction-processing workload without downtime is hard because writes must remain consistent, and a split-brain situation can make reconciliation difficult.
Agent-ready development
Schemas and company workflows.
The founders discuss declarative schemas: a file representation of the desired database structure can give agents a more direct view than reconstructing the current state from a long migration history. Migration execution and review still matter.
They describe hiring contributors to work on Postgres itself and using the company’s asynchronous documentation as context for internal AI workflows. The discussion closes on measuring work completed rather than treating token spending as the goal.
More from this channel
1:46:40EN
Latent Space· Oct 27, 2026Podcast · 3 speakersInside Anthropic’s Agent Infrastructure and Product EngineeringGuest Katelyn Lesse, Angela Jiang · Host swyxKatelyn Lesse and Angela Jiang discuss managed agents, evaluations and sandbox design.
16:34ENLatent Space· Oct 26, 2026Podcast · 3 speakersIan Quigley on Testing Scientific Agents in Premeasured WorldsGuest Ian QuigleyLeash Bio’s Minos environments replay physical measurements so agents can explore medicinal chemistry decisions at scale.
1:25:33ENLatent Space· Oct 22, 2026Podcast · 2 speakersWhy GPU Capacity Comes in Three-Year Rectangles — Erik Bernhardsson, ModalGuest Erik Bernhardsson · Host Vibhu SapraModal’s CEO tells Latent Space why AI demand almost never looks like a three-year GPU contract, how the company pools compute across thousands of customers, and why sandboxes grew from a product they almost killed into a third of revenue.
26:41ENLatent Space· Oct 22, 2026Podcast · 2 speakersAnekha Sokhal on Moshi, a Recommendation Engine for Commodity TradersGuest Anekha Sokhal · Host swyxThe Moshi founder tells swyx how she would rebuild a trader’s workflow: connect market data, private files and conflicting sources, then push the second-order effects people would otherwise miss. The hard part is not the dashboard. It is getting someone to actually use it.
43:33ENLatent Space· Oct 21, 2026Podcast · 3 speakersInside Tenex: AI-Native Engineering, Incentives and Enterprise AdoptionGuest Alex Lieberman, CJ Hess · Host swyxAlex Lieberman and CJ Hess describe their engineering model.
36:43ENLatent Space· Oct 20, 2026Podcast · 2 speakersEvan Conrad on GPU Markets and Delivering ComputeGuest Evan Conrad · Host swyxSF Compute’s founder discusses contracts, infrastructure and market incentives.