Live: selve.me
Status: Production (Deployed Dec 19, 2024)
SELVE is a personality assessment platform measuring 8 scientifically-validated dimensions. Built on 2M+ psychological test responses with an average reliability of α = 0.861 (Excellent).
- Frontend (Next.js 16): Assessment UI, results dashboard, authentication
- Backend (FastAPI): Assessment engine, scoring algorithm, PostgreSQL database
- Notebooks (Jupyter): Data validation, psychometric analysis, dimension extraction
- Data (2M+ responses): Big Five, HEXACO, 16PF datasets from Open Psychometrics
| Dimension | Measures | Reliability |
|---|---|---|
| LUMEN ✨ | Social Energy | α = 0.897 |
| AETHER 🌫️ | Emotional Stability | α = 0.872 |
| ORPHEUS 🎵 | Empathy | α = 0.837 |
| ORIN 🧭 | Organization | α = 0.821 |
| LYRA 🦋 | Openness | α = 0.804 |
| VARA ⚖️ | Honesty | α = 0.900 |
| CHRONOS ⏳ | Patience | α = 0.937 ⭐ |
| KAEL 🔥 | Assertiveness | α = 0.821 |
Average: α = 0.861 (Excellent by psychology standards)
- Redis is required for backend progress tracking. See Redis Setup Guide
cd frontend
pnpm install
pnpm dev # http://localhost:3000# Start Redis first
docker compose up redis -d # or: brew services start redis
cd backend
source .venv/bin/activate
uvicorn main:app --reload # http://localhost:8000
# or: make devcd notebooks
jupyter lab
# Open: 01_big5_validation.ipynb (start here)Frontend: Next.js 16, TailwindCSS, Clerk Auth
Backend: FastAPI, Prisma ORM, Neon PostgreSQL
Deployment: Vercel (frontend), AWS EC2 (backend)
Data Science: Python, pandas, numpy, Cronbach's alpha analysis
- ✅ 8-dimension personality assessment
- ✅ Clerk authentication (Google, LinkedIn OAuth)
- ✅ Personalized results dashboard
- ✅ Anonymous assessments (no login required)
- ✅ Mobile-responsive UI
- ✅ Friend/family assessment invitations
- 🚧 Adaptive testing (temporarily disabled)
All 8 dimensions validated across 2,086,821 responses:
- Big Five IPIP: 1,015,341 responses (5 dimensions)
- HEXACO-60: 22,786 responses (VARA, CHRONOS)
- 16PF: 49,159 responses (KAEL)
Methodology documented in /notebooks/README.md.
Frontend: Auto-deploy via Vercel (push to master branch)
Backend: AWS EC2 instance with FastAPI + Uvicorn
Database: Neon PostgreSQL (serverless)
cd backend
# Install dependencies
make install
# Start server
make dev # http://localhost:8000
# Run tests
make testcd frontend
pnpm install
pnpm dev # http://localhost:3000