Reflections on building internal tools after AI changed the workflow
I know this isn’t the usual kind of letter you get, but it felt like a good moment to pause and reflect.
Ever since AI entered the app development space in full swing, things have changed quickly. I’ve watched customers’ PMs prototype in tools like Lovable or Replit, then hand things off to engineers using Cursor to build internal apps. It mostly works. There are loose ends, but people seem happy.
The flip side is that not all new internal apps are being built on my low-code platform anymore.
So over the past three months, my team and I have been building two things on DronaHQ.
First, a vibe-code tool for internal software. It takes a different approach from Cursor, Replit, or Retool—focused on letting teams build production-ready internal apps quickly, without losing structure or clarity.
Second, an AI agent builder—an orchestrator that makes it possible to build RAG, chat, voice, and autonomous agents without writing code.
The important part is that these tools interoperate. You can move between them easily and build real systems, not disconnected demos.
What I’m wishing for isn’t hype or virality. I’m hoping for posts that reach the right people—the ones who read carefully, try things out, and decide whether it actually solves a problem they have.
Specifically: • A Show HN that gets a fair look • Comments that challenge the ideas, not the intent • Traffic driven by curiosity, not outrage • A few users who stick around and push the product forward
I’ll do my part: clear docs, honest positioning, and steady shipping.
If a bit of good timing or luck helps along the way, I’ll take it.
Thanks for rooting for builders, even the ones shipping close to the holidays.
— A builder