New Relic’s Radar technology uses artificial intelligence to order and present alerts and error messages to admins monitoring server infrastructure. (New Relic Photo)

New Relic is building some new artificial intelligence capabilities into its core infrastructure monitoring product that will allow operations engineers to browse a curated feed of alerts and updates deemed most worthy of their attention.

The services are called New Relic Applied Intelligence (NRAI), and the company plans to unveil them later today at its Future Stack conference in New York City. They’ll become part of New Relic’s service, which allows engineers and managers responsible for the health of complicated infrastructure to get up-to-the-second updates on application performance, server issues, or networking glitches.

Radar is perhaps the most interesting of the three new services. It “is going through all of your systems data and finding everything that is useful and actionable, and presenting them in a personal feed,” said Nadya Duke Boone, director of product management for platform at New Relic. The company showed off Radar — previously known as Seymour — at its Future Stack conference last fall, but the technology is now generally available.

Ops engineers often face a never-ending stream of information from monitoring products like New Relic that can be difficult to parse or prioritize without help, and that’s what Radar aims to provide. One trial customer using the product discovered a key server running out of disk space that hadn’t been checked in a while, and Radar was able to detect that it was nearing capacity and send an alert, possibly preventing a more serious problem down the road from a server that looked fine at a glance.

“That’s a problem we can solve. It makes you feel safe,” Duke Boone said, detecting problems not just with newly deployed servers or code but with systems that otherwise appear to be running normally.

New Relic, which is headquartered in San Francisco but has several hundred engineers based in Portland, will also introduce two other AI-driven monitoring features Tuesday.

New artificial intelligence features from New Relic let admins set specific baseline parameters for triggering alerts. (New Relic Photo)

Customers will now be able to set baseline alerts using the New Relic Query Language (NRQL), which lets users write scripts to ping the monitoring database for specific information. “Anything you can write a query for that returns a number, we can set an alert for what’s expected,” she said, allowing customers to automate the process.

New Relic is also adding artificial intelligence capabilities to its application-performance monitoring product, which allows ops teams to understand how custom applications are performing on their infrastructure.

New Relic engineers using the product internally often found themselves comparing sets of errors, looking for common patterns between different sets of errors to help them understand what just happened to the application. Turns out, computers are way better and faster at that then people, and this feature will be able to automatically compare errors with historic data and normal performance metrics, helping them figure out what went wrong this time.

You’ll need to be a Pro customer of New Relic’s services to take advantage of Radar and the new NRQL alerts, but all customers will have access to the error monitoring service. New Relic’s revenue has increased fairly steadily since its 2014 IPO, with a 37 percent jump year-over-year in its most recent quarter and full-year revenue growth of 31 percent, although it has yet to achieve profitability.

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.