Supabase Bets on Postgres to Build the Next Platform for Developers

Paul Copplestone didn’t plan on raising capital, certainly not while tucked away in Wānaka, a remote town on New Zealand’s South Island better known for postcard views than venture capital dealmaking. But venture capital came anyway.

Accel partner Gonzalo Mocorrea didn’t send a calendar invite. He asked for an address.

Then, in a scene that sounds more like a startup legend than a standard Series D, Mocorrea showed up in person. “He literally showed up on my doorstep in Wānaka, which is really not easy to get to,” Copplestone recalled. Over the next couple of days, the pair would have sporadic chats “a couple hours here and there” with no formal meeting on the books. Then Mocorrea called in backup: Accel partner Arun Mathew.

Copplestone, both amused and alarmed, tried to dissuade him. “‘Oh no, don’t come! We haven’t agreed to anything!’” he said. Mathew didn’t listen. After more than 24 hours of travel, including two flights and a long drive, he arrived in Queenstown. They had dinner, caught up the next morning, and by the end of the visit, Accel offered a term sheet.

That term sheet turned into Supabase’s $200 million Series D round, now valuing the startup at $2 billion. The round, led by Accel, also brought in Coatue, Y Combinator, Craft Ventures, and Felicis, along with individual investors like Kevin Weil (OpenAI), Guillermo Rauch (Vercel), and Taylor Otwell (Laravel).

Mathew said the face-to-face was essential. “I needed to sit across the table, look him in the eye, and really believe he’s going to do something else,” he told Fortune. “That’s necessary, certainly at this valuation. We know what greatness looks like, we believe that.”

Reluctant Raise

Copplestone insists Supabase didn’t need the capital. Nor were they actively raising. “We had to shake Paul down in his hometown of Wānaka to get his attention,” Mathew said. But the opportunity aligned with a momentum that was already impossible to ignore.

Founded in 2020, Supabase describes itself as a backend-as-a-service platform, a cloud infrastructure stack built around Postgres, offering integrated services like Auth, Storage, Realtime messaging, and Edge Functions. Think Stripe for payments or Vercel for frontend deployment, Supabase is the equivalent for backend infrastructure.

That positioning has won over more than 2 million developers who now manage over 3.5 million databases on the platform. The company sees more than 10,000 new databases launched every day. And all of this, remarkably, has been achieved with little to no traditional marketing growth has been fueled almost entirely by word of mouth and community engagement.

It’s also fueled by what insiders are calling the rise of the “vibe coder” , a new class of developers using platforms like Bolt, Cursor, Lovable, and Vercel’s v0 to build fast, often AI-driven applications. “Our sign-up rate just doubled in the past three months because of vibe coding,” Copplestone said. “Bolt, Lovable, Cursor, all those.”

For Mathew, Supabase’s success aligns with a broader thesis. “In every major platform shift, there’s always value created at the database layer,” he said. He points to Oracle and MongoDB as proof. While databases are littered with failed ventures, the ones that succeed tend to create lasting value. Supabase’s bet is that the next enduring platform will be built not on proprietary engines, but on a rock-solid, developer-first approach to Postgres.

That wasn’t always the obvious move. When Supabase launched in 2020, Postgres’s position wasn’t as secure as it is today. But with AI, the rise of semi-structured data, and features like pgvector Supabase doubled down. To date, the company has contributed more than 25 open-source tools to the ecosystem and sponsors several community-led Postgres projects.

Built on Moneyball Thinking

Copplestone, a third-time founder whose past companies include Nimbus for Work and ServisHero, describes Supabase as a product of the pandemic and his own hard-earned lessons. “I went through playing startup the first time, where you raise the money and put the posters up. You pay top dollar, you hire, tell people how many employees you’ve got and then, of course, we ran out of money.”

This time, he’s doing things differently. Supabase’s team is remote by design and spread across the world from Peru to Macedonia. “It’s a bit like Moneyball,” said Copplestone. “We found these really good humans, and they’re not necessarily in San Francisco. What matters to us is that they’re extremely competent, but also low-ego, good people.”

Roughly 28% of the team are former founders themselves. And while Supabase lacks an HQ, it hasn’t let that stop in-person interactions. The company runs global launch weeks, releasing one new feature per day and organizing meetups in more than 100 cities. Community events have now taken place in 43 countries.

Supabase’s slogan “build in a weekend, scale to millions” isn’t hyperbole. The company cites multiple businesses that started with its $25 Pro plan and scaled to over a million users without migrating their backend.

Even the company’s name Supabase is a story. Copplestone had been searching for a domain that evoked “superlative base,” but nothing was available. “I could only find S-U-P-A,” he said. The unexpected bonus? “It reminded me of the Nicki Minaj song ‘Super Bass,’ so I chose that name so I could send memes to my cofounder Ant [Wilson], who’s very good at memes.”

If it sounds irreverent, that’s the point. Supabase wasn’t engineered to look impressive. It was engineered to work fast, well, and with the ergonomics developers actually want. Rather than bolting on features, the team designed core infrastructure pieces like Auth and Realtime as deeply integrated but composable tools, with clean APIs and a unified UI.

This design philosophy is beginning to define the company’s moat. Supabase has become the backend of choice for the new generation of no-code and AI-augmented tools. And their developer-focused approach is winning loyalty in one of software’s most competitive segments.

In December, the company launched an AI Assistant that won Product Hunt’s Data Product of the Year. The same month, user activity reached record highs. Now, the team is betting that the broader ecosystem of builders from indie devs to enterprise will grow up with them.

One upcoming release includes a native plugin for LangChain, enabling developers to build and deploy RAG (retrieval augmented generation) pipelines in minutes. Another is a low-latency vector search engine optimized for hybrid workloads, built on top of pgvector and tuned for large-scale AI applications.

Whether that will continue to work at a $2 billion valuation remains to be seen. But if Mathew’s 24-hour trek and the investors around the cap table are any indication, a lot of people believe it will.

They just had to go to Wānaka to prove it.