Your AI guide to achieving your North Star
Transform one ambitious goal into 64 tangible actions using the Ohtani/Harada Method. Track progress with an AI coach that adapts.
Polaris is an AI-powered goal coaching app built for the Encode Club "Commit To Change" Hackathon.
The Problem: 92% of resolutions fail because they're vague, lack a system, and have no accountability.
The Solution: Polaris uses the same methodology that helped Shohei Ohtani become a baseball legend—breaking one big goal into 8 supporting pillars and 64 specific daily actions, with AI coaching to keep you on track.
- AI-Guided Goal Crafting — Conversational wizard to define your goal, pillars, and actions
- Visual Mandala — 9x9 heat map showing your progress across all 64 actions
- Natural Language Check-ins — Just tell Polaris what you did, it maps to your actions automatically
- Adaptive Coaching — AI notices patterns, celebrates streaks, nudges cold pillars
- Framework: Next.js 15 (App Router, React 19)
- Database: Convex (real-time subscriptions)
- Auth: Clerk
- AI: Claude Sonnet 4.5 via Vercel AI SDK v6
- Styling: Tailwind CSS v4 + shadcn/ui
# Clone the repo
git clone https://github.com/0xSardius/polaris.git
cd polaris
# Install dependencies
npm install
# Copy environment variables
cp .env.example .env.local
# Fill in your API keys in .env.local
# Initialize Convex
npx convex dev
# Run the development server
npm run devOpen http://localhost:3000 to see the app.
polaris/
├── src/
│ ├── app/ # Next.js App Router pages
│ ├── components/ # React components
│ │ ├── ui/ # shadcn/ui components
│ │ ├── mandala/ # Mandala grid components
│ │ ├── chat/ # Chat interface
│ │ └── craft/ # Goal crafting wizard
│ ├── lib/
│ │ ├── ai/ # AI prompts and logic
│ │ ├── opik/ # Observability setup
│ │ └── utils.ts # Utility functions
│ ├── hooks/ # Custom React hooks
│ └── types/ # TypeScript types
├── convex/ # Convex backend
│ └── schema.ts # Database schema
└── docs/
└── PRD.md # Product Requirements Document
Color Palette: Night Sky + Polaris Gold
- Background: Deep space (
#0a0a1a) - Primary: Polaris gold (
#fbbf24) - Heat states: Cold → Warming → Warm → Hot → Fire
Targeting:
- Best Use of Opik ($5,000) — Full observability of AI coaching effectiveness
- Personal Growth & Learning ($5,000) — Harada Method implementation
Created by Takashi Harada and famously used by Shohei Ohtani:
- 1 Goal — Specific, measurable, time-bound
- 8 Pillars — Balanced supporting areas (not just skills)
- 64 Actions — Daily behaviors, not outcomes
Ohtani's high school chart included "keep room clean" and "be caring" alongside baseball skills. Balance matters.
This is a hackathon project, but feedback and ideas are welcome!
MIT
Built with ☕ for the Encode Club "Commit To Change" Hackathon
Powered by Claude, Vercel AI SDK, Convex, and Opik