Skip to content

About Agent Monitor

Agent Monitor allows you to track and analyze your agents' performance across multiple runs. It provides a comprehensive, time-based view of agent activities, helping you optimize your AI operations.

Key Capabilities

  • Performance Tracking: Monitor your agents across multiple runs, gaining valuable insights into response times, execution patterns, and overall efficiency.
  • Dual-View Analytics:
    • All Runs: Comprehensive data on all agent instances, offering a holistic view of endpoint interactions.
    • Model Runs: Focused analytics on Gen AI node executions, providing deeper insights into language model performance.
  • Detailed Metrics: Access critical performance indicators, including total runs, average response times (P90 and P99), and failure rates, allowing for precise performance tuning.
  • API Key Usage Monitoring: Track and manage the utilization of API keys across your agent ecosystem.
  • Gen AI Node Analysis: Gain specific insights into the performance of individual Gen AI nodes within your agents.
  • Advanced Filtering and Search: Use time-based searches and custom filters to drill down into specific scenarios or periods.
  • Detailed Run Information: Access comprehensive logs and debug information for each agent run, facilitating efficient troubleshooting and optimization.

Best Practices

  • Regularly monitor the Agent Monitor dashboard to track performance trends.
  • Use time-based searches to analyze specific periods of interest.
  • Combine multiple filters to further analyze particular scenarios.
  • Review detailed run information to understand the flow of data through your agent and identify any potential issues.

By effectively leveraging Agent Monitor, you can ensure the optimal performance of your agents and make data-driven decisions to improve your operations.