The studio

Small studio. Senior people.

Twenty-four people across six time zones, all of whom have shipped native apps before yours. No offshore delivery pod, no bench, no learning on your budget.

Founded

2016

Team

24 people

Why we exist

We got tired of good ideas shipping as bad apps.

Devora started in 2016 because two platform engineers kept being handed the same job: rescue an app that had been built by people who had never really used the platform they were building for.

The pattern was always similar. A team that knew the web had been asked to produce something for a phone, and had produced something that technically worked. It launched, it got two stars, and the fix cost more than the original build. The problem was almost never effort — it was that mobile has its own physics, and you cannot reason your way to them from a browser.

So we built a studio that only does this. Swift and Kotlin, iOS and Android, with the design and backend work that a real app needs sitting around that centre rather than competing with it. Ten years later the shape has not changed much, which we take as a reasonable sign it was the right shape.

We are deliberately small. Four to six senior people on a project, one team from pitch to launch, and a hard limit on how many engagements run at once. It means we turn work away. It also means the person who promised you something is the person who has to deliver it, which turns out to matter more than almost anything else.

Apps shipped to the stores
47Apps shipped to the stores
Downloads across our portfolio
12.4MDownloads across our portfolio
Average post-launch store rating
4.8Average post-launch store rating
Median crash-free session rate
99.97%Median crash-free session rate
How we operate

Six things we don't negotiate.

Native, or an honest no

We build in Swift and Kotlin because that is what the hard problems require. When a project genuinely doesn't need native, we say so — and lose the work.

The build is the status report

You get a real build on a real device every week. Nothing focuses a schedule like an app you can open.

Performance is a constraint, not a phase

Frame budgets, cold-start limits and crash gates run in CI from week one. Performance work you defer becomes a rewrite.

Accessibility ships with version one

Dynamic Type, VoiceOver and contrast are part of the definition of done. Retrofitting them costs several times more and is always worse.

You own everything

The code, the design system, the documentation, from the first commit. No proprietary framework, no licence, no exit fee.

Small teams, senior people

Four to six people who have shipped this before, not fourteen who are learning on your budget.

Ten years

The short version.

  1. 2016Founded in San Francisco by two ex-platform engineers.
  2. 2018First App Store feature. Team reaches nine people.
  3. 2020Went remote-first. Opened a second base in Lisbon.
  4. 2022Crossed five million cumulative downloads.
  5. 2024Delivered our first SOC 2 Type II mobile engagement.
  6. 2026Twenty-four people across six time zones. 47 apps shipped.