|
Hello Reader, Our beta launch for Sucana failed. And the feedback was brutal. Beta users said, the application is,
Now imagine this: weeks of late nights, coding until 4 AM, finally hitting “launch”… and the first thing you hear from users is exactly that. It stings. Then my cofounders Victor & Virgil called me and said, “Let’s flip it all around.” And I’m sitting here thinking,
You see, neither of my cofounders can code. But they’re full of “brilliant” ideas. lol But, here’s the thing, though,
Our strategy from day one was clear:
Because we knew feedback would shape Sucana far better than we could ourselves. So instead of spending months polishing a “perfect” product in isolation, we built the core features, shipped in 3 months, and went straight to users. The brutal feedback gave us the clarity we needed: Not what we thought the product should be, but what agencies actually want. That’s why our beta ‘failure’ feels like a win. The LessonYour first version isn’t supposed to be perfect. Now we’re rebuilding Sucana’s core flow, faster, smoother, designed for the way agencies really work. If you’re building something:
Thanks for following along on the journey. I’ll keep sharing both the wins and the painful lessons as we build. - Vinod P.S. I’ve been sharing even more of the behind-the-scenes journey on YouTube. Subscribe here to follow along. |
I build AI systems to help founders and enterprises acquire and support their users. Currently supporting 4M+ active app users. Back to code in my 50s after 12 years in management.
Last week, I showed how to build a personal context folder for Claude Code from scratch. That approach is useful because you understand exactly how the system works. But not everyone wants to spend an hour deciding which files, folders, and instructions to create. So I packaged the starting point into a free Claude Code Context and Skills Kit. The setup is simple: Open an empty folder in VS Code. Copy one prompt. Run it with Claude Code. Claude will interview you about your: Background and...
Your AI forgets you every time you open a new chat Every new session starts the same way. You explain your stack. Your project. Your brand. What you decided last week. Then you switch tools and explain it again. The fix is not a better prompt or longer context memory. It is a folder that remembers, sitting on your machine, that every AI tool you use can read. I showed my own brain folder in the last post. It has been taking shape for months, and it runs everything I build. Several people...
Let me show you what actually happens when Virgil "just asks Nelly." Nelly is the AI brain behind Virgil's launches. I built her. Virgil talks to her in plain language over Telegram. For example, he'll ask Nelly to analyse the last three launch campaigns. "Hey, analyse these emails and tell me which one actually made the money." Simple, right? Looks like magic. It is not. It is plumbing. Here is what fires behind that one line. The brain is connected to Kit, so it can read every email that...