JOSEPH GABRIEL LAGONSIN
Information Editor
New Relic has launched New Relic Autopilot and New Relic Floor Reality, geared toward companies constructing or utilizing AI brokers.
The 2 choices symbolize the following stage of its platform technique round what it calls agentic AI-first operations. One is designed to deal with incident response duties as an automatic web site reliability engineering agent, whereas the opposite feeds observability information into AI brokers that clients already use.
Autopilot is positioned as an out-of-the-box automated SRE software. It begins analysing an incident as quickly as an alert fires, with the intention of serving to groups establish root causes, assess whether or not motion is protected and resolve what to do subsequent.
The service contains specialist brokers for areas reminiscent of Kubernetes, together with instruments for Kafka troubleshooting and cross-stack root-cause evaluation. It additionally attracts on an organisation’s runbooks and retrospectives by New Relic Information, whereas exterior Mannequin Context Protocol connections to Jira and GitHub pull in code particulars.
The product additionally contains long-term reminiscence to seize inside operational information and make it out there throughout groups. It may be used by a brand new agent-focused interface within the New Relic platform, in Slack and thru automated workflow actions triggered by occasions reminiscent of a extreme incident, a software program deployment or a scheduled process.
Two approaches
Floor Reality takes a distinct method. Relatively than offering a managed agent, it provides clients’ current AI brokers entry to New Relic’s observability information by instruments designed for machine use slightly than dashboard-based interplay.
The product works with third-party and in-house brokers, together with GitHub Copilot, Claude Code, AWS DevOps and customized orchestrators. New Relic says it’s supposed to reveal richer insights than can be found by public APIs or primary question instruments, whereas decreasing the variety of software calls an agent wants to search out helpful info.
New Relic additionally cited what it described as buyer proof for Floor Reality, saying a big enterprise self-measured a 1.1% error fee throughout greater than 1,300 customers. It didn’t establish the shopper.
The launch displays a broader shift in how software program distributors are adapting monitoring and operations instruments for AI-driven workflows. In conventional observability merchandise, engineers often examine dashboards and alerts immediately. AI brokers, against this, want information entry by interfaces that may be queried programmatically and acted on routinely.
Camden Swita, Head of AI at New Relic, outlined that view in feedback launched with the announcement.
“Operations are going headless. AI brokers will not log in to view dashboards. They’re going to pull what they want by APIs, cause about it, and act,” Swita mentioned.
“That is precisely what New Relic Floor Reality delivers: it grounds your individual brokers in the actual reality of your techniques, by APIs as an alternative of UIs. And for groups who’d slightly we run the agent, there’s New Relic Autopilot, our skilled agent operated for you. Each sit on the identical wealthy information substrate, and both manner the toil is lowered,” he mentioned.
Operational focus
New Relic’s emphasis is on incident administration and operational response, the place AI instruments are more and more being utilized to cut back repetitive work for engineering groups. The tough a part of an incident typically lies not in detecting a failure, however in understanding its trigger and deciding whether or not a remediation step is protected, particularly in massive and sophisticated software program environments.
By combining telemetry, inside documentation and hyperlinks to code administration techniques, New Relic is in search of to position its observability platform nearer to the centre of automated operations. That method additionally permits it to serve two teams of consumers: those who desire a vendor-managed agent and those who desire to construct or orchestrate their very own.
The cut up is important as a result of many massive organisations are experimenting with a number of AI brokers throughout software program growth, operations and help. Instruments that may plug into these current techniques could also be simpler to undertake than merchandise that require groups to exchange their most popular platforms.
New Relic has framed each choices round a shared information layer. In follow, which means the identical underlying observability info can both be utilized by New Relic’s automated agent or accessed by customer-run brokers by devoted interfaces.
New Relic says the results of its automated evaluation is offered as a structured account grounded in system information and operational context.