• Make sure your phone and your computer are on the **same WiFi network**. • Open your computer's terminal and type `ipconfig` (Windows) or `ifconfig` (Mac) to find your IPv4 Address (e.g., `192.168.1.x`). • Open `src/app/index.tsx` and `src/app/apikey.tsx` in your code editor and change `SERVER_URL = 'http://localhost:8000'` to your IP address. • On your computer, run `npm run start` inside the `mobile` folder. • Download the **Expo Go** app from the Google Play Store on your Android phone. • Scan the QR code in your computer terminal with your Android camera or Expo app.

This app currently uses a "Long-Context" workaround for 100% accuracy on a single book. However, to scale to hundreds of PDFs (your long-term R&D goal), we must use **Advanced RAG**. **Advanced RAG Architecture consists of:** • **Metadata Tagging:** Extracting hard data (planets, houses, chapters) during ingestion into the Vector Database. • **Agentic Query Rewriting:** Using a small AI to intercept the user query, deduce planetary relationships (e.g. "Cancer is 2nd from Gemini"), and rewrite it into a strict SQL-style database filter. • **Cross-Encoder Re-Ranking:** Pulling 100 potential rules from the database and using a lightning-fast evaluator AI to score and perfectly rank the top 5 most relevant rules before generating the final answer.