New Relic announced an open‑source feature called New Relic AI Coding Observability that extends production‑grade monitoring into the AI‑assisted coding phase. The tool is aimed at engineering and platform leaders who need real‑time visibility, cost control, and security governance for AI coding assistants such as Claude Code, Cursor, and GitHub Copilot. As AI‑driven code generation accelerates, many organizations are finding that these assistants operate outside traditional observability stacks, creating “blind spots” that can hide performance problems, unexpected spend, or compliance violations. By bringing the same rigor that New Relic applies to production services to the earlier development stage, the new feature promises to turn fragmented, unmonitored AI usage into a measurable, auditable enterprise advantage.
New Relic Announces AI Coding Observability Feature
The Intelligent Observability Company said the new feature will close “blind spots” that appear when AI coding assistants operate outside traditional observability stacks. Chief Product Officer Brian Emerson emphasized that organizations “can’t manage what you can’t see,” noting that rapid adoption of AI assistants is scaling risk alongside output. New Relic positions the feature as a unified, vendor‑neutral pane of glass that normalizes telemetry from multiple AI coding tools and correlates it with existing production infrastructure. The announcement highlights that engineering teams will move away from “blind trust” in their assistants and instead gain concrete insight into how the tools behave while code is being written, tested, and prepared for deployment.
Technical Design and Supported Tools
New Relic AI Coding Observability is built on open standards, including the OpenTelemetry protocol and the Model Context Protocol (MCP). By leveraging OpenTelemetry, the solution can ingest traces, metrics, and logs from any supported assistant without requiring proprietary agents. MCP adds a layer that captures model‑level context—such as prompts, token usage, and inference outcomes—so that telemetry reflects both the developer’s intent and the AI’s response.
The feature runs queries in a “local‑only / zero‑outbound” mode, ensuring that data never leaves the user’s private network. This design choice supports data sovereignty, privacy, and regulatory compliance, which are especially critical for enterprises handling sensitive code or operating in tightly regulated industries.
Supported assistants include Claude Code, Cursor, GitHub Copilot, Windsurf, and Amazon Q, providing broad coverage across the fragmented AI coding landscape. By exposing telemetry from these tools, the solution promises to replace anecdotal productivity claims with measurable data, surface hidden failure modes, and enable cost forecasting against budgets. Because the code is both open‑source and source‑available, security teams can audit the implementation, verify privacy safeguards, and even extend the telemetry model to additional assistants as the market evolves.
Enterprise Relevance and Pricing
The announcement cites Gartner’s prediction that 90 % of enterprise software engineers will use AI code assistants by 2028, underscoring the growing financial impact of these tools. New Relic’s solution aims to turn AI‑related spend from an “unmonitored expense” into a trackable line item, with alerts that trigger when budgets approach predefined thresholds. In addition to cost control, the feature offers productivity metrics that capture time‑saved, code‑quality improvements, and detection of inefficiencies such as repeated regeneration attempts.
The feature will be released on June 23 at no additional cost, subject to New Relic’s standard ingest rates. A local‑only mode will be added shortly after the initial launch, giving early adopters the option to run the observability pipeline entirely within their own data center or private cloud.
Key Takeaways
- New Relic AI Coding Observability will be available as an open‑source feature on June 23, with no extra licensing fee beyond standard ingest rates.
- The tool supports five AI coding assistants—Claude Code, Cursor, GitHub Copilot, Windsurf, and Amazon Q—and uses OpenTelemetry and MCP for vendor‑neutral telemetry.
- It offers cost‑tracking, security‑focused “local‑only” operation, and productivity metrics to replace anecdotal reporting with hard data.
TechInsyte's Take
The addition of observability to the AI coding stage gives CIOs and engineering leaders a concrete mechanism to monitor spend, security, and performance before code reaches production. While the open‑source release removes licensing barriers, adoption will depend on how quickly organizations can integrate the new telemetry into existing monitoring pipelines. Buyers should watch for early‑stage feedback on the local‑only mode and the practical usefulness of the cost‑alerting features.
Source: Businesswire
