README_Deployment.md

🚀 Deployment Guide: DocuMind AI

Your system is now Production-Ready and Scalable.

🏁 Quick Start (Production Mode)

1. Install Dependencies:

   pip install waitress

2. Start the Server: Instead of python app.py, run:

   python serve.py

   (This uses 6 threads vs 1, handling multiple users smoothly)

3. Start the Background Worker: Keep your existing worker running:

   python watcher.py

   (Or start_all.bat if you updated it to use serve.py)

---

🛡️ Key Features Enabled

Feature	Function	Benefit
WAL Mode	Database Concurrency	Uploads don't block Chat. Chat doesn't block Uploads.
RBAC Caching	Permission Optimization	Access control checks are 100x faster.
Strict RAG	Anti-Hallucination	Answers ONLY from documents. "I don't know" if unsure.
Auto-Classification	Smart Sorting	Crypto docs -> Technology. Contracts -> Legal.
Hybrid Retrieval	Access+Search	Students can find "Symmetric Cipher" docs they have access to.

🔮 Future Maintenance

• Files: All sorted files are in data/sorted. If you need to re-scan everything, move files from sorted back to incoming.
• Database: data/users.db. Backup this file regularly.
• Scripts: Maintenance scripts (like DB fixes) are now in scripts/maintenance/.
• Logs: Check logs/app.log for access audits and errors.

📈 Next Levels (Optional)

If you grow to 10,000+ users, refer to docs/analysis/POSTGRES_MIGRATION_ANALYSIS.md to switch the engine.