
I wrapped up my last company earlier this year and thought I would take a proper break. I wanted more time with my 3 year old son and I also considered taking a stable job for a while.
But it kept hitting me that right now might be the best moment in my career to build something new. The tools feel different. The speed is different. The opportunity feels wide open. My co-founder Han felt the same. We both knew we would regret it if we did not try.
So after two months of idea exploration, we started building PlayGraph AI (playgraph.ai).
Ten years ago, even building a simple MVP took months of grinding. Today it feels completely different.
With tools like Cursor and Claude Code, a single technical founder can move like an entire early stage team. With PlayGraph AI, Han and I often joke that our current velocity would have required 10 to 12 engineers in the old model. Now it is just the two of us. We estimate we are operating at a 5 to 7x multiplier, and some teams I know run at 10x. We are still learning.
AI native building is especially good for new projects because you are not stuck with old architecture or years of tech debt. The tricky part is making sure that you have proper engineering process in place so that you do not accumulate new debt by shipping too fast. It's a different topic.
This is the part that excited me most.
Learning new skills used to be a slow and painful process. Blog posts, YouTube tutorials, trial and error. If Google did not have the answer, you were stuck.
Now, if you can describe the problem clearly and include a screenshot, AI gets you 80 to 90 percent of the way. The last part is judgment and testing.
For example, I became fairly competent in Figma in a short amount of time. Auto layout, components, variants, references. These used to take months for me.
AI did not replace learning. It removed the wasted time. This makes it easier to build a well rounded founding team because you can pick up new skills quickly.
In my earlier founder days we used pitch decks to test demand. The problem is that slides never convert users.
Now you can build a real landing page in a few hours and let people interact with something that feels real. It mirrors an actual customer journey. They see it, click it, explore it, and decide for themselves.
This gave us much more honest signal when we were testing PlayGraph AI in the first two months. We just put the page out and watched how people behaved.
Because of these efficiency gains, the entire model changes.
You can focus on revenue per employee from day one.
You can stay small for a long time.
You do not need to raise capital to build.
For many founders, this will be the preferred way to build. Small teams, fast loops, closer to users, faster to learn.
I will keep sharing our progress as we build PlayGraph AI. If you want to follow along or chat, feel free to message me. Always happy to meet other builders.