Things we believe

Agents aren't replacing humans.

They're tools. Powerful ones. But still tools. The power comes in harnessing them to enable human judgement, taste, and delight.

Building is getting close to not being a blocker at all.

Software used to be opaque to everyone who couldn't write it. That's changing fast. The distance between an idea and a working system is the shortest it has ever been, and gjalla is here to bridge the gap between the tools and the end goals.

Trust is built on understanding.

Speed without understanding is how teams lose control of their own systems. We need better ways to keep what the code does legible to the people who are responsible for it.

Why I'm building gjalla

I've spent my entire career building trust, reliability, and seamlessness into technologies that most people never think about. The best infrastructure fades into the background. It gives people a stable place to build, solve the problems they care about, and fit into the world the way they most want to.

I love platform work, and I really believe in it's importance. People should be able to confidently build without needing to understand every layer beneath them first.

What changed?

AI agents. So innovative and fun to tinker with, and so much potential to build with. They introduced an entirely new layer that we could not see, inspect, or yet have much reason to trust.

The internet works because decades of engineering, protocols, and shared standards make an extraordinarily complex system feel dependable. Agents do not have that foundation yet. Their instructions, context, decisions, corrections, and spend are scattered across tools and sessions. Most teams cannot answer basic questions about what their agents are doing, why they are doing it, or whether all that effort is producing useful work.

This problem is getting bigger as adoption grows. Teams are moving faster, spending more, and feeling more like things are out of their control.

gjalla exists to solve this problem.

We're bringing agents' hidden layer to the surface so that we can actually use it to our advantage instead of being beholden to it.

Although part of platform work is optimization... making sure that tooling works seamlessly, performantly, and without worry... you can't do any of that without first understanding it. gjalla provides a place where teams can see what their agents worked on, where the tokens went, which context shaped the work. Only then can you identify where time is being lost to confusion, rework, or repeated mistakes.

What have we learned?

I've learned a lot over the last nine months as a founder. We've had the chance to work alongside thoughtful early customers and users while the entire industry learns together. The shape of our solution has changed as the tools and needs around us have changed. The vision has not.

We've seen teams go from no coding agents to running some workflows autonomously. We've seen them run into new problems are agents deliver hundreds of PRs per day. We've seen an entirely new vocabulary come into view as our old assumptions are completely disrupted.

What has stayed consistent? The need for visibility. Understanding. Confidence. Control.

We want teams to spend less time reconstructing what their agents did, correcting the same mistakes, or wondering whether their spend produced anything valuable. We want them focused on the work only they can do: making good decisions, applying judgement and taste, and building things that matter.

That is why we're building gjalla. The tools will keep changing. We're building the place that helps teams understand them, get the most from them, and stay in control.