Back in December I posted on LinkedIn that I was at a crossroads with dicehub. The platform wasn't getting the traction I expected. Most of our usage concentrated around the Pedestrian Wind Comfort template, which worked great, but overall growth had flatlined. I kept asking myself whether it was the product, the marketing, the market timing, or something else entirely.
I laid out the three options I was sitting with. Double down with more features, more templates, more marketing, and go all in. Keep dicehub running in maintenance mode for our existing users while I shifted focus to something new. Or pull the plug completely and move on. I wrote it honestly because I genuinely didn't know which way to go, and I asked for advice.
The response
It blew me away. Dozens of people reached out with advice, ideas, collaboration offers, and honest perspectives. Some of you even came by our office in Mainz to talk it through in person. If I somehow missed replying, I owe you a thank you too. That message and the conversations around it changed the way I was looking at the problem.
Where things stand
Short answer: dicehub is very much alive.
The longer answer is that three things shifted between December and now.
The feedback was overwhelmingly positive. The people closest to what we build weren't telling me to stop. They were telling me the opposite. That alone was enough to keep me going.
The customers who use dicehub are happy to keep using it. This is the signal I care about the most. It's easy to get lost in growth curves and adoption dashboards and forget that the people already on the platform are the best judges of whether you've built something worth keeping.
The new generation of LLMs changed what I can do as an engineer. The recent Opus and GPT releases, combined with dicehub's codebase (over half a million lines of code, a mature microservice structure, clear guidelines) turned out to be a great match. Work that used to take me months now takes days. Sometimes hours. Last month alone I closed over two hundred issues. That number still feels unreal to type out.
A small army of agents
What this practically means is that I'm actively developing dicehub again, and I'm doing it with a small army of agents running all day. New features, long-standing bugs, infrastructure work that had been sitting in the backlog for quarters; all of it is finally moving. The product direction feels re-energized for the first time in a long while.
None of this is a victory lap. The original questions from December, about traction and market fit and whether this is the right hill to climb, are still there. But the answer to "should dicehub exist?" has quietly become yes, and that matters.
To everyone who wrote back, visited the office, or quietly rooted for us: thank you. You kept this thing alive.
Sign up and run your first simulation in dicehub.