DIMWOS — Train your brain like an LLM.
Inspiration
I noticed that the process AI labs use to train models — start broad, then build judgment, then tool use, then compress what you've learned, then stress-test yourself — is basically a really good blueprint for how humans grow too. I'd already written it out as a 10-stage framework. This felt like the right excuse to turn it into an actual app instead of just a doc.
What it does
You type in a goal. The app breaks it into stages (based on that 10-stage framework) instead of just answering questions like a chatbot. Each stage points to real resources — books, videos, courses — instead of AI-generated text. Stage 1 is free. Unlocking the rest of the plan requires a subscription, handled through RevenueCat.
How I built it
This is a concept I'm just starting on for the Shipathon, not a finished build yet. The plan: a React Native / Expo app, RevenueCat for the subscription and paywall, and a simple system that maps a user's goal onto the 10 stages I already have written out.
Challenges I ran into
Turning a long written framework into something that works as a phone screen, not a document, without it feeling like homework. Also figuring out where exactly the free/paid line should sit — generous enough to show value, but a real reason to subscribe.
Accomplishments that I'm proud of
The full 10-stage framework is already mapped out and written. I've got the core concept and tagline locked in, and a clear paywall model (free stage 1, paid stages 2–10) figured out before writing a line of code.
What I learned
Structuring the RevenueCat paywall around "stages" instead of a flat free/paid split is a simpler model than I expected — it maps naturally onto progress instead of just access.
What's next for DIMWOS
Build the MVP: goal input, a working stage 1 experience, RevenueCat paywall wired up for stages 2–10, and a curated resource list for each stage.
Log in or sign up for Devpost to join the conversation.