Join Our Communities
IN
IN
JPN
EN
From the TZ APAC EGG Web3 Incubator to the TZ APAC Tezos Incubator and eventually to Fortify Labs, we spent several years in the room with founders. Not advising from a distance, but genuinely in it with them. Present when things unraveled and occasionally, when something clicked.
What follows is what lessons and insights we learnt from it, and why it eventually led us to wanting to build something of our own.
Across the founders we worked with, we observed a pattern that emerged.
The teams that struggled most were rarely short on talent or commitment. They were short on clarity. Clarity about who they were building for, clarity about what problem they were actually solving, and perhaps most importantly, clarity about when to stop chasing something that was not working.
The interesting part is that this was not a knowledge problem. Most founders knew, somewhere, that they had drifted from the original problem, but the context was scattered. Decisions made three months ago had lost their reasoning and conversations that shaped the direction of the product lived only in the heads of the people who were in the room. Notes, when they were taken at all, disappeared into folders no one returned to.
By the time a team realized they had been solving the wrong thing, they had already spent months on it.
We saw this enough times that it stopped feeling like individual mistakes and started feeling like a structural problem, a problem worth sitting with.
Three things, more than anything else, stayed with us.
The first is that speed has very little to do with moving fast. The teams that moved with the most velocity were not the most aggressive ones, they were the most aligned. Everyone knew the current thinking, the current priority, the reasoning behind the current direction. That kind of alignment does not happen by accident but requires a kind of infrastructure that most early teams did not build.
The second is that cutting is a skill that almost no one develops naturally. Staying on a problem past the point where the signals have turned, because stopping feels like failure. What we learned, slowly and sometimes the hard way, is that the ability to cut cleanly and move forward is one of the most genuinely valuable things a team can cultivate. It is also one of the hardest to teach.
The third is that institutional memory is fragile. This is obvious in retrospect but it surprised us in practice. A team of six people, eighteen months into a product, can lose enormous amounts of context through a single departure, a long break, or simply the accumulation of too many meetings with no clear record of what was decided and why. It just slows everything down until someone notices the team keeps revisiting the same questions.
These were not observations we made from a comfortable distance. They were things we lived alongside the founders we worked with, and often, experienced ourselves.
Running Fortify Labs meant constant context switching, a high volume of conversations and decisions scattered across many threads at once. We felt the cost of lost context directly. We built workarounds and different systems but none of them were good enough.
At some point it became clear that the product we needed did not exist in a form that actually solved the problem. So we decided to build it.
That is where Keystone Lab came from. A direct response to something we kept running into and could not find a good answer to anywhere else.
Keystone Lab is our product studio. We embody the same discipline we applied to working with founders: sharp problem definition, fast iteration, cutting without attachment, we now apply to our own work.
Our first product in progress is Panora. It is a privacy-first AI meeting assistant with voice diarisation that lets you query across your recorded conversations, with the option to keep your data completely local. Ask it a question and get an answer: who said what, when, in which meeting, and in what context.
It is, in the most direct sense, our answer to the problem we described above. Context that lives inside conversations and then disappears. Decisions that lose their reasoning between one meeting and the next. Panora is built to be the memory that never fades.
If the problem resonates with you, follow Panora on X. If where we are headed resonates with you, follow Keystone Lab on X or visit the website. The best conversations we had always started with someone saying this sounds familiar, we want to hear from you.
