Observability: Enhance monitoring features

[vprashrex]

Is your feature request related to a problem?
The current system lacks observability features, resulting in an inability to track API calls, latency, and evaluate workflows effectively. This hampers debugging and performance assessment.

Describe the solution you'd like
Enhance the system with observability features to monitor usage and performance:

• Monitor API calls per endpoint and NGO
• Track latency (total, internal, and LLM provider)
• Capture end-to-end eval latency
• Enable basic request-level tracing

Original issue

Describe the current behavior
A clear description of how it currently works and what the limitations are.

• No tracking of API calls per endpoint or NGO
• No clear latency breakdown (internal vs LLM provider)
• No end-to-end latency tracking for eval workflows

Describe the enhancement you'd like
A clear and concise description of the improvement you want to see.
Add observability to track usage and performance.

• Monitor API calls per endpoint and NGO
• Track latency (total, internal, and LLM provider)
• Capture end-to-end eval latency
• Enable basic request-level tracing

Why is this enhancement needed?
Explain the benefits (e.g., performance, usability, maintainability, scalability).

• Helps debug issues and identify bottlenecks
• Improves system reliability and performance
• Provides visibility into usage and scaling

Additional context
Add any other context, metrics, screenshots, or examples about the enhancement here.