Real-Time Telemetry in Development: Using MCP and Datadog for AI-Assisted Coding

Description

Whitney demonstrates an innovative approach to AI-assisted development by integrating real-time telemetry data through the Datadog MCP server with Claude Code. In this video, she demonstrates how using OpenTelemetry instrumentation during development (not just production) can give AI agents accurate, real-world data about system behavior. She showcases Commit Story, a demo application that automatically generates journal entries from Git commits and Claude Code conversations. The real magic happens when debugging - using telemetry data to catch AI agent bad assumptions before they lead to suboptimal solutions. We see compelling examples of "time-travel debugging" where comparing data flow from weeks ago to current behavior helps identify root causes like bypassed filter functions. 00:00 Introduction and Commit Story demo 02:23 How the journal generation system works 03:13 Exploring trace IDs and telemetry instrumentation 04:18 Requesting system diagram from Claude Code 05:42 Debugging timeout issue with telemetry data 07:04 Time-travel debugging - comparing past vs present 08:08 Claude Code's complete system understanding 09:40 Using telemetry to find existing functionality 11:37 Challenge of keeping codebases clean with AI 12:28 Telemetry as a time machine for debugging 13:07 Example of heavily instrumented helper function 14:52 Auto-instrumentation agent workflow 15:21 OpenTelemetry Weaver and semantic conventions 17:28 Agent steps: correlation and validation 18:39 Black boxes in software development 19:15 Using telemetry to inform development choices #Telemetry #OpenTelemetry #Datadog #MCP #ClaudeCode #AIAssisted #Observability #Debugging #DevTools #SoftwareDevelopment #MCPServer #Instrumentation #AIAgents #DeveloperTools #CodeQuality #TimeTravel #DistributedTracing #Logs #Metrics #Traces