A Personal Letter on Kypso's Journey

I'm thrilled to share that Kypso has been acquired by Incredibuild. This is an exciting moment for us, and I wanted to take a moment to reflect on the journey that brought us here - a journey rooted in a deeply personal mission that Tomer and I took on together.

Tomer and I had a shared frustration from our time at Snyk: watching engineering teams consistently underperform when we knew they had the potential to achieve so much more. It wasn’t theoretical. We felt it every day. And we found ourselves asking: What would it look like if developers could focus on building software - and nothing else?

With that ideal in mind, we started Kypso. When we started, we made a deliberate choice to bootstrap. No investors, no external pressure, just us and our own capital. Looking back, this decision shaped everything about Kypso. Every dollar had to justify itself. Every feature had to prove its worth immediately (lesson that took us some time to learn).

We shipped our first version fast - it worked, but it wasn’t perfect. We kept iterating, listening, adjusting, and by April 2024, something clicked. We signed our first paying customers. I remember those moments more than any metric - each one felt like a quiet but profound “yes, keep going”. It meant we were solving something real.

From the very beginning, we understood something that feels obvious now but wasn't then: context is everything. While everyone else was racing to wrap LLMs in a nice interface, we built a knowledge graph and we worked hard on predictability. We knew that without understanding your codebase, your team's patterns, your organizational knowledge, AI was just expensive guessing. So we invested heavily in capturing that context and improving users’ trust, making it the foundation of everything we built.

Looking back at it, we've learned countless lessons along the way. But two stand out - two challenges that ultimately shaped our decision to join Incredibuild.

First, we realized we were solving a second-order problem. Developer productivity unrelated to code is critical, but for most organizations, it wasn't the fire they were fighting today. It was the problem they would fight after they solved their coding bottlenecks or their immediate infrastructure issues.

Second, we built an opinionated product. We did provide building blocks, but we built high-level abstractions designed to be simple to use and maintain. The problem was that every new customer came with slightly different requirements, different workflows they wanted to automate. Products like n8n and Gumloop were gaining traction because their lower-level building blocks could adapt to whatever problem came next. We optimized for simplicity when the market needed adaptability.

We had happy customers who loved what we built, but we weren't seeing the explosive growth that signals true product-market fit. The gap between customer satisfaction and market traction kept us searching for answers.

Here's the thing about hard lessons - you often learn them too late. We kept bootstrapping, doubling down on serving our customers and building technology that worked. Then we tried something different - a code review agent. It was off our initial focus because it was purely code-focused, but the market response was immediate. Customers got it instantly. It solved a first-order problem. But we didn’t feel it was something that aligned with our first principles.

That realization led us to a decision - we could pivot and start over with these learnings, or we could find a partner where our technology and insights could make an immediate impact.

Incredibuild brings the massive customer base and distribution that we needed. Our technology is a perfect strategic fit to Incredibuild’s vision and together, we're set to make an even greater impact on developer productivity worldwide.

Tomer and I learned that being right isn't enough. Timing matters. Distribution matters. Sometimes the best way forward is joining forces with a team that can help you reach the impact you've always wanted to have.

We're incredibly excited about this next chapter. The problem we set out to solve - helping engineering teams reach their full potential - is still the mission. We just have a much better context to solve it in now.