Production-ready Python LLM proxy handling tenant-level cost attribution, rate limits, and fallback routing via OpenTelemetry.

What’s included:

• Tenant-Level OpenTelemetry (OTel) Cost Attribution
• Dynamic Hard Limits & Token Rate Limiting
• Multi-Provider Fallback Routing (OpenAI, Anthropic, Local)
• Plug-and-Play Streamlit Cost Dashboard & Docker Compose