Training Personal Data

Turn your wearable data into actionable health insights.

This project syncs Oura Ring and Wahoo data into PostgreSQL and generates Weekly Health Insights using SQL + GPT. It stores normalized metrics and structured AI analysis so you can query trends, build dashboards, and automate feedback loops.

What it does

• Sync Oura Ring (activity, sleep, readiness, heart rate, workouts, tags)
• Sync Wahoo workouts
• Normalize and store daily metrics (+ raw JSON)
• Compute Monday→Sunday weekly insights
• Generate, parse, and persist AI analysis (text + JSONB)
• Automate weekly insights via GitHub Actions

Quick start

1. Set environment variables in a .env file (DB + Oura + OpenAI)
2. Sync data for a date range:

bb run:oura "2024-01-01" "2024-12-31"

3. Generate weekly insights (pass any date within the week):

bb -m training-personal-data.insights.week 2024-12-30

3. Sync Wahoo workouts (requires Wahoo OAuth client + tokens):

bb run:wahoo "2024-09-01" "2024-09-30"

4. Run tests:

bb test

Learn more

See the full project guide with architecture, workflows, schemas, troubleshooting, and conventions:

• AGENT Guide: AGENT.md