New Relic launches a platform for AI agents and new OpenTelemetry tools

When companies begin rolling out AI-agents at scale, an inevitable question emerges: who watches the watchers? New Relic, one of the key players in the observability market, has decided to answer it first — and presented both a platform for creating and managing AI-agents, as well as a suite of tools for deeper integration of OpenTelemetry data.

To understand the significance of this announcement, you need to look at what's happening in corporate AI right now. 2025 has become the year of agents — autonomous AI systems that don't just respond to queries, but independently execute chains of tasks: analyzing data, calling APIs, making decisions. But the more complex the agent, the harder it is to understand what exactly it's doing and why. Classical monitoring tools, honed for microservices and containers, are poorly suited for this. Agents aren't just code — they're sequences of reasoning, model calls, tool invocations, any of which can go wrong.

This is precisely the gap that New Relic is targeting. The new AI-agent platform allows enterprises not only to create and deploy agents, but to observe their entire lifecycle. This involves tracing chains of calls, monitoring latency of language model invocations, tracking token costs, and detecting anomalies in agent behavior. Essentially, New Relic is transposing the APM — Application Performance Monitoring — philosophy to the world of agentic AI, where instead of tracking HTTP requests you need to monitor prompts, and instead of database response time you need to track inference latency.

The second part of the announcement is no less important, though it sounds more technical. New Relic is expanding its integration with OpenTelemetry — an open standard for collecting telemetry data that has become the de-facto lingua franca of the observability world in recent years. OTel allows you to collect metrics, logs, and traces from any systems in a unified format without being tied to a specific vendor.

New Relic's enhanced tools for working with OTel data mean that companies will be able to combine information about the performance of traditional infrastructure and AI-agents in a single dashboard. This is critically important: an agent that calls an external API with a three-second delay might look "broken" at the AI level, when the problem actually lies at the network or backend level.

For the market, this move by New Relic is a signal of the formation of a new segment: observability for AI. Until now, monitoring of AI-systems has remained fragmented. Developers used LangSmith for debugging LangChain chains, Weights and Biases for model experiments, and for production monitoring often made do with custom solutions. New Relic offers a unified point of entry — and does so based on an open standard, which lowers the barrier to entry and the risk of vendor lock-in.

Competition in this space will intensify rapidly. Datadog is already experimenting with AI-monitoring, Dynatrace is developing its causal analysis tools, and startups like Arize AI and Langfuse are targeting the exact same niche. But New Relic has a scale advantage: thousands of enterprise customers who already use the platform for infrastructure monitoring and can add AI-observability without switching their stack.

There's a broader context as well. As AI-agents take on increasingly responsible tasks — from managing cloud infrastructure to processing customer requests with real financial consequences — the question of their transparency becomes not just technical, but regulatory. The European AI Act and similar initiatives require explainability and auditability of automated decisions. Observability tools for AI-agents are not a luxury, but necessary infrastructure for compliance with future regulations.

New Relic is betting that the era of agentic AI will require the same mature monitoring infrastructure that we're accustomed to seeing in the world of cloud applications. If this bet is correct, the company finds itself in the right place at the right time. If the rollout of agents slows down or the market fragments into dozens of incompatible frameworks, the task of consolidating all data streams becomes significantly more complex. But one thing is clear now: AI observability is ceasing to be a niche topic and is becoming a mandatory layer of the corporate stack.