Our customers aren’t coming to us asking what can you do with AI. Our customers are saying my environment’s incredibly complex. There’s so much going on, so many microservices all trying to work in concert to deliver a great customer experience, so many possible points of failure from the customer experience to the application to the infrastructure.” —Lew Cirne, Founder and CEO, New Relic

Instead of wondering what they can do with AI, New Relic customers want intelligent solutions to real problems. They want AI solutions that can learn from their data patterns and suss out hidden issues so they can proactively prevent outages, downtime, and performance problems.

That is exactly what we hope to provide with New Relic Applied Intelligence (NRAI), a new set of features in the New Relic Digital Intelligence Platform designed to analyze your data to give you actionable results. New Relic Applied Intelligence powers features like Dynamic Baseline Alerts, Error Profiles, and Radar (formerly known as Project Seymour), which works to surface patterns and offer suggestions as to the possible root cause and how to address it.

When it comes to AI, we’re committed to staying away from the hype and concentrating on solving real customer problems. But Radar is only part of the NRAI initiative, which includes many smart capabilities built into the New Relic Digital Intelligence Platform (more on that below).

Radar helps you see through your data fog

Project Seymour was rolled out to limited customers earlier this year. But today Seymour ditches its code name in favor of a new moniker, Radar, that better describes its place in the New Relic Digital Intelligence Platform.

Radar is a destination page for you to access advice and recommendations on issues or abnormalities found in your environment. Applying statistics, algorithms based on the distillation of hundreds of engineer-years of experience, and machine learning, Radar is built to cut through the fog of data to help you see important things that can be hard for humans to detect on their own. Additionally, Radar learns from your feedback to provide a personalized feed containing issues, abnormalities, and recommendations that might be applicable to your role and interests:

radar personalized feed

Personalized feed (containing advice, anomalies, and potential issues) provided by New Relic Radar [click to enlarge]

The Radar feed consists of different kinds of advice, perspective, and events information called “cards.” Each Radar card contains not only the information on the event/abnormality/potential issue, but also the possible root cause. Many Radar cards use New Relic’s patent-pending root-cause-analysis technique to figure out what might be causing the issue. And because we realize the importance of collaboration to quickly resolve issues, Radar comes with a built-in Slack integration so you can easily share its recommendations with the right people on your team.

new relic radar card

Collaborate to resolve issues found by Radar faster, using built-in Slack integration.

But that’s not all. As part of our focus on helping customers as they move to the cloud, Radar has added new analysis for Amazon Web Services (AWS) users, including cards that identify Amazon Relational Database Service (RDS) connection leaks, that highlight regional performance issues, and that help to find underutilized hosts and make recommendations to save on hosting costs—we know almost everyone is interested in that one!

Underutilized AWS resources

Underutilized AWS resources detected by Radar [click to enlarge]

Amazing early feedback

We have received amazing feedback from our early customers of Radar. Richard Petersen, senior consultant for IT applications at Nationwide Insurance, says, “New Relic Radar has surfaced response time delays in our applications that our team wasn’t aware of and helped us improve performance. For example, Radar identified a week-over-week performance drift and enabled our team to fix it before it made it to production, eliminating any potential impact on customer experience.”

James Nguyen, an engineering manager at Riot Games, had this to say: “Riot Games has a wide array of applications, but it’s hard to fully absorb and prioritize actions from all the data we gain from monitoring. An automated tool like New Relic Radar has been valuable in guiding us to issues for services that operate in the dark. For example, Radar proactively directed us to a disk space utilization issue for a machine we don’t pay much attention to, enabling us to add disk space before it impacted users.”

Another customer received a N+1 query antipattern advice card in their Radar feed. The company’s solutions architect hadn’t been aware of the inefficiencies in the code and immediately shared the card with their engineering team to get it fixed.

New Relic Applied Intelligence gives you views you can use

Over the last few years, we have added a number of smart services to the New Relic Digital Intelligence Platform. These services analyze your full-stack data monitored by New Relic agents in your environment to provide smart features like Dynamic Baseline Alerts, Error Profiles, and Radar. NRAI is the collection of these intelligent services, and we plan to continue to grow NRAI to help solve your real-world problems.

A great example is the Baselines Service, which includes a built-in ensemble of algorithms automatically applied to your data to determine which one is the best fit. With the best algorithm (chosen based on your data), Baselines predict the behavior of your application with minute-scale granularity. Baselines power our Dynamic Baseline Alerts and NRQL Baseline Alerts capabilities, which help customers create dynamic alerts based on historical data. (See Dynamic Baseline Alerts Now Automatically Find the Best Algorithm for You for more information on how this works.)

ensemble of algorithms chart

This ensemble of algorithms are automatically evaluated against your data to find the best fit.

This Baselines Service also powers anomaly detection in Radar. This platform demonstrates how we plan to bring AI into more and more parts of the New Relic product lineup.

Another smart service in NRAI that recently went live is New Relic APM Error Profiles, which uses statistics to analyze error attributes and present collections of error attributes that have significantly different traits than do non-errors. Using New Relic APM Error Profiles, you can quickly figure out patterns seen in error attributes and significantly reduce MTTR. Now that this feature is generally available, go give it a try!

new relic apm error profiles chart

New Relic APM Error Profiles shows you the patterns found in your error attributes [click to enlarge]

New Relic’s unique advantage

As our customers’ environments grow increasingly complex, they not only need full-stack visibility but also automated algorithms to analyze the extensive amount of data created by their deployments. The available algorithms also keep on evolving and newer, better algorithms are introduced frequently.

New Relic Radar and NRAI can leverage the New Relic agents that instrument your entire application stack to analyze this full-stack data. And because New Relic is a pure SaaS platform, we can include newer algorithms in Radar and NRAI as they become available, without requiring customers to take any additional steps. (On-premise solutions often make customers download and deploy upgrade patches with every new update.)

Where to get more information

For more information on Radar and NRAI, check out these links to the New Relic Blog and New Relic Documentation:

And if you have any questions, feel free to ask the experts at the New Relic Online Technical Community.


Chhavi Nijhawan is a senior product marketing manager at New Relic. She is passionate about new technologies. View posts by .

Interested in writing for New Relic Blog? Send us a pitch!