Live

Dreambook

An AI dream interpreter, built end-to-end through an AI pipeline — zero lines of code written by hand.

Visit live ↗

Quick facts

Launched
2026-01
Stack
Flutter · FastAPI · LLM pipeline · S3
Audience
Consumer, global
Status
Live

Why I built it

Dreambook was the testing ground. I wanted to find out what an AI-driven pipeline could actually carry — architecture, content, ASO, monetization — with one person making decisions and the AI doing the execution. So I picked a real consumer product and shipped it to the stores to find out where the method holds and where it breaks.

The premise was simple and a little uncomfortable: if AI can handle the execution layer, what does a solo founder actually do? My answer, after building this: you make decisions, set quality bars, and debug the places where the AI confidently goes in the wrong direction. That’s the job. Not writing code. Not producing content. Deciding, evaluating, correcting.

Everything in Dreambook was built through that pipeline. The Flutter app, the FastAPI backend, the LLM interpretation chain, the S3 content engine, the ASO copy, the onboarding flow, the monetization setup. One person, end to end, shipped to both stores and first paying customers.

What I learned

  • Spent four months blaming users for a drop-off that was my onboarding. The fix was obvious once I actually watched a session replay. → Breakdown
  • Rewrote store titles and watched CTR move from 0.4% to 6%. The lesson: ASO copy is a product decision, not a marketing task.
  • Built an autonomous content engine — script to final cut, no manual step in the loop. This is the part that doesn't look possible until you see it run.
  • Regional pricing and per-language voices: where 'global' defaults go wrong, and how granular you actually need to get.

Proof

Screenshots and metrics — coming in Phase 6.

Deep dives

Engineering breakdowns from building Dreambook.

Want to see the product itself?

Visit live ↗ ← All projects