Production deployment guide

[artugro]

Context

No documentation on how to deploy Intuno to production. Self-hosting is one of Intuno's key differentiators (Ollama embeddings, no OpenAI requirement), but without deployment docs, nobody will actually self-host it.

Proposed: docs/DEPLOYMENT.md

Sections

1. Prerequisites — Docker, PostgreSQL 15+, Redis 7+, Qdrant 1.7+
2. Docker Compose production setup — production-ready compose file with proper resource limits, restart policies, health checks
3. Environment variables checklist — every required/optional env var with descriptions and example values
4. Database setup — PostgreSQL with pgvector extension, connection pooling recommendations
5. Vector database — Qdrant deployment options (Docker, Qdrant Cloud), collection sizing
6. Redis — caching and rate limiting configuration
7. Reverse proxy — nginx/Caddy configuration with SSL termination, WebSocket support for /ws
8. Health check monitoring — using the /health endpoint, recommended alerting
9. Scaling — horizontal scaling guidance (stateless app servers, shared DB/Redis/Qdrant)
10. Embedding provider setup — OpenAI vs Ollama configuration for air-gapped deployments