New Relic brings production performance data directly into GitHub

The gap between code and production

Software engineering teams are increasingly using AI tools like GitHub Copilot to speed up the development process. While these tools help write code faster, there is a significant disconnect between the quality of code during a review and how it actually performs in a live environment.

According to New Relic's 2026 State of AI Coding, 94% of leaders believe AI-generated code is high quality during the review phase. However, 78% of those same leaders reported experiencing more incidents once that code was actually deployed to production.

This gap often leads to prolonged triage times when performance problems occur. Engineers frequently have to leave their primary tools to search for the root cause of an issue, which slows down the ability to fix bugs and maintain stable software.

Headless observability in GitHub

New Relic is addressing this by bringing 'headless' observability directly into GitHub workflows. This means developers can see real-time application performance data without having to log into a separate dashboard or switch between different software platforms.

The New Relic Change Tracking for GitHub Copilot integration connects production telemetry with the deployment workflow. It allows engineers to see the impact of new code immediately, helping them understand how human and AI-generated changes affect downstream services.

By keeping this information in the same place where the code is built, the system helps teams stay focused. It aims to turn faster coding into reliable delivery by providing the necessary context to make informed decisions about which code is ready for production.

Automated checks and incident response

The new integration includes several automated safety checks to ensure software stability. It can run pre-merge checks against a baseline, act as a canary gate after a deployment, or perform regression detection to catch issues early.

If a regression is detected, the system automatically routes the issue to GitHub Issues and Copilot for a fix. It also sends updates to Slack to keep the rest of the team informed about the status of the problem.

Additionally, New Relic Autopilot for GitHub MCP Server provides context for incident response. This agent can draw on a company's specific codebase to triage alerts, identify the root cause of a failure, and recommend specific fixes to the developers.

Monitoring the AI agents themselves

Beyond monitoring the final software, New Relic is also providing tools to monitor the AI agents used during the coding process. The New Relic Preflight integration is an open-source capability designed to help developers understand how they are interacting with coding agents.

This tool provides visibility into model efficiency, latency, and token consumption directly within the local integrated development environment (IDE) or command line interface (CLI). It helps teams see errors during execution and identify issues with how they are using AI tools.

By providing this local telemetry, engineering teams can optimize their costs and evaluate the accuracy of different large language models. It gives them a unified view of both the development process and the production results of the code they create.

Availability and next steps

New Relic is rolling these features out in different stages of availability. New Relic Change Tracking for GitHub Copilot is currently available in public preview for users to test.

The New Relic Autopilot for GitHub MCP Server and New Relic Preflight for GitHub Copilot are already generally available. These tools are intended to help teams maintain quality as AI continues to accelerate the pace of software development.

The company's broader strategy is to move away from reactive firefighting. By using AI-driven automation and integrated telemetry, they want to help businesses protect revenue and optimize technology spend in real-time.