Mindle

The story

Mindle was the biggest and longest project I've ever worked on. It was crazy how demanding it was and how much time it took me and my co-founder Gabriel to set it up.

Making a text editor alone is not easy. It has a lot of caveats, and many of them are hard to solve with LLMs, even hard to describe to one.

What I learned about startups

The biggest lesson, and the reason I dropped the project, is that I fell in love with a toxic idea. I wasn't solving a problem, and that showed when people didn't pay for the app.

Second lesson. A second brain looks appealing to a lot of people. What could be better than a system that saves everything in your life? In short, nothing can substitute your mind. The creativity of connecting ideas on the spot is unique to us humans, and a memory system can't replace it.

Third lesson. Do a lot more interviews, and tell your customers what you're building as early as you can. That way you don't waste time thinking about and building features that seem easy but still eat your time. Customers guide you toward what they need. Don't confuse them with friends and family, because they will probably tell you your app looks cool.

What I learned building it

I learned a lot on the technical side too, from a cache system for photos to shareable notes with public links, how to make an MCP and how to connect to WhatsApp.

The most important one was building a RAG system to manage a big amount of information. Thanks to Gabriel I learned a lot about the subject, and I'm using similar ideas in adjacent projects.