AI Streaks is a lightweight product for turning one good AI plan into a real 7-day execution streak.
The idea is simple: people already use ChatGPT or other assistants to generate learning plans, workout plans, interview prep plans, and side-project roadmaps. The missing layer is follow-through. This app adds structure, onboarding, persistence, and a day-by-day learning loop so the plan becomes something a user can actually complete.
- Helps a user define a learning goal
- Generates a structured 7-day plan
- Stores the plan and progress in Supabase
- Walks the user through a daily lesson flow
- Uses an LLM as a tutoring layer during execution, not just ideation
This project is a good example of the kind of product work Logan likes to build:
- take a familiar AI behavior
- identify the missing operational layer
- wrap it in a clear product flow
- make the outcome more useful than a single chat session
It is less about "AI magic" and more about workflow design, accountability, and shipping practical interfaces around model behavior.
The current implementation includes:
- Landing page explaining the problem with AI-generated plans that never get used
- Authentication and onboarding flow with Supabase
- User dashboard for managing templates and active learning paths
- 7-day template generation flow
- Guided day-by-day tutoring experience tied to a stored plan
- Next.js
- TypeScript
- Supabase Auth and Database
- Tailwind CSS
- Framer Motion
- OpenAI API
- Install dependencies
npm install- Create
.env.local
NEXT_PUBLIC_SUPABASE_URL=your_supabase_project_url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key-
Apply the Supabase SQL and migrations in
supabase/ -
Start the app
npm run devThen open http://localhost:3000.
- The repo name is
ai-streaks, but the broader concept is "turn AI plans into progress." - Some internal naming still reflects an earlier prototype. The public framing here reflects the actual product direction more accurately.