AI Summary
5 min readWhy Agents Love Supabase
Supabase co-founders Paul Copplestone (CEO) and Ant Wilson (CTO) sat down with host swyx at their annual Select conference to explain how the company evolved from a real-time Postgres startup into a $10 billion infrastructure layer that has become the default backend for the AI coding boom. The conversation reveals a company that accidentally timed itself perfectly: built for developers over six years, then discovered that everything they'd built was exactly what AI agents needed.
The Three Innovations: Performance, Scale Up, Scale Out
Supabase announced three major database innovations at Select, each targeting a different growth vector. The first, Oriole DB, is a drop-in replacement for Postgres's storage engine that runs 1.5 to 2x faster and eliminates background tasks like vacuum and bloat. After a year of rigorous testing—PG regress, TPCC benchmarks, Jepsen—they released it in beta under Apache 2 license, including a patent they acquired and are now giving away for free.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Latent Space: The AI Engineer Podcast
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:01) **Host Intro: Why Supabase Matters for AI Engineers** - The host explains his close relationship with Supabase as an early angel investor and why the company is worth studying for AI builders, framing the episode around Supabase's momentum as the standard for hosting Postgres.
- 2 (05:28) **Opening Announcements: Supabase Compute, Oriole DB, and Multigress** - Co-founders Copple and Ant join to reveal the three major technical announcements from their customer conference, starting with a new long-running compute product.
- 3 (12:37) **Scale Out: The Terso Acquisition and the Billions-of-Databases Problem** - Supabase announces the acquisition of Terso to solve the challenge of launching billions of tiny databases per month, enabling a new "prototype to petabyte" pathway.
- 4 (16:15) **The Vibe Coding Wave and Chapter Three of AI at Supabase** - The founders describe how the AI boom has created three distinct chapters for their business, with the most recent seeing an explosion in sign-ups and conversion rates.
- 5 (23:04) **Origin Story: The Accidental Hacker News Launch** - The co-founders recount how Supabase went from eight users to eight hundred overnight when a random user launched them on Hacker News with the tagline "Open Source Firebase Alternative."
- 6 (28:43) **From "Build in the Weekends" to "Prototype to Petabyte"** - The founders discuss how the scale of modern AI-powered startups has changed, with new projects monetizing within days rather than months.
- 7 (33:38) **How Supabase Became the Default Recommendation of Every Frontier Model** - The founders explain the accidental timing and deliberate community-building that led to Supabase being the top recommendation from every major LLM for an integrated backend.
+ Full timestamped outline available in the app
Show Notes
AI agents are creating more software—and that changes what a backend needs to do. Supabase co-founders Paul Copplestone and Ant Wilson join swyx to discuss running long-lived workloads alongside the database, making schemas easier for agents to underst...