We’ve always had a low tolerance for things that don’t work properly.
Once, on a family holiday in Switzerland, the hotel Wi-Fi didn’t work and it quickly became less of an inconvenience and more of a problem to solve.
Instead of putting up with it, Tobie bought some tools and rewired the entire hotel network – we should add that we did get permission to do this and it took a whole week to do.
It’s not something he would have left half-done either – if something doesn’t work and it’s worth fixing, he’ll keep going until it does. That’s how we’ve both always approached things.
Over a decade ago, while running a golf analytics startup, we found ourselves managing four separate databases, time-series, document, graph, and relational, just to make one product function.
It quickly became clear that we were spending more time dealing with infrastructure than actually building anything useful.
This didn’t make much sense to us and our intolerance for badly designed, outdated, and overly complex systems meant we weren’t particularly willing to accept that as the status quo.
It wasn’t just about how things worked, but how they felt to use as well, something that, for Jaime, goes back to an early fascination with Apple products and a long-standing interest in good design.
So, in a move that probably felt excessive at the time, we decided to build a new kind of database that could do everything in one place.
That became SurrealDB. And pretty quickly, we realised that the same mindset applied beyond the product because if we were going to try to build something differently, it didn’t make much sense to build the team in a conventional way either.
We approached hiring in much the same way and questioned the default assumptions and focused on what actually mattered.
Rather than focusing on traditional credentials or expected career paths, we looked for curiosity, adaptability, and people who were comfortable figuring things out as they went.
Over time, that led us to build a team that includes opera singers, a former Broadway performer, and several self-taught developers.
In practice, that mix has been one of our biggest advantages.
People with artistic backgrounds often bring strong communication skills, discipline, and the ability to perform under pressure, while those without formal technical training tend to approach problems without the usual assumptions.
The result is a team that is more willing to question how things are done, which is essential when you’re trying to rethink them entirely.
A similar pattern played out in how the company itself grew. We didn’t spend much time trying to build visibility in the usual places, so we weren’t particularly active on Reddit, Discord or Twitter, and we didn’t rely on heavy marketing.
Instead, SurrealDB spread more quietly, through conversations in shared workspaces, at conferences, and through people discovering it on their own.
There were multiple occasions where we found ourselves talking to someone and only later realised they were already using SurrealDB internally.
It’s not a conventional way to build a developer product, but it reinforced something we’ve come back to repeatedly, that if you solve a real problem, people will find you.
That has been reflected in the way the product has scaled. Developer interest in SurrealDB was immediate.
It became the fastest-growing database in history, reaching over 10,000 GitHub stars in just four weeks.
What business wants from politicians: truth, stability and support
Today, it has surpassed 30,000 stars and more than 2 million downloads, with developers using it in over 190 countries, from individual builders to large enterprises.
We took much the same approach when it came to investment. While we were approached by a number of well-known venture firms early on, we turned down several offers because they didn’t feel like the right fit.
We were wary of following a standard playbook of rapid hiring and aggressive scaling before the product had reached the level of maturity we were aiming for.
Instead, we focused on finding partners who understood what we were building and were comfortable with a longer-term approach.
That approach has since led to SurrealDB raising $44m (£33m) in total funding, including a $23m Series A extension announced in February 2026, with investors including Chalfen Ventures, Begin Capital, FirstMark and Georgian.
Unconventional
Looking back, none of this has followed a particularly conventional path.
SurrealDB started as an internal tool, shaped around the needs of client projects, and the initial build was done privately before it was unveiled.
During that time, Tobie focused on the deep engineering work, while Jaime handled the operational side of the business, from client handling and recruitment to partner communication and the general day-to-day logistics of keeping things moving.
If there’s one thing building SurrealDB has reinforced for us, it’s that we’ve never been particularly comfortable accepting things just because that’s how they’re usually done, and more often than not, that’s been worth questioning.
That said, taking that route isn’t the easy option. It takes a lot of time, a lot of work, and a genuine interest in what you’re doing to stick with it.