Observability

orq.ai offers comprehensive Observability capabilities for monitoring your AI applications and deployments.

Core Observability Features

Built-in Monitoring

Once your Deployments are in use, orq.ai automatically provides:

  • Logs - Record requests, responses, errors, and performance metrics
  • Traces - Track end-to-end request flows across services and operations
  • Threads - Monitor concurrent execution contexts and conversation flows

OpenTelemetry Integration

For advanced observability, orq.ai supports industry-standard OpenTelemetry for detailed application monitoring:

  • Observability Frameworks - Quick setup guide for sending OpenTelemetry traces to orq.ai
  • OTLP Endpoint: https://api.orq.ai/v2/otel - Standard OpenTelemetry Protocol endpoint
  • Multi-Language Support - Python, Node.js, Java, and other OpenTelemetry-supported languages

OpenTelemetry Use Cases

Application Performance Monitoring

  • Track LLM request latencies and token usage
  • Monitor embedding generation and vector operations
  • Analyze RAG pipeline performance

AI Agent Observability

  • Trace multi-step agent workflows and tool usage
  • Monitor conversation flows and context management
  • Debug complex agentic interactions

Custom Instrumentation

  • Add spans for business logic and custom operations
  • Track model performance across different environments
  • Implement distributed tracing across microservices