New Relic today added more than 15 additional capabilities to its observability platform, including retrieval augmented generation (RAG) tools, which are all enabled to one degree or another by artificial intelligence (AI).
Announced at a New Relic Now+ conference, these additions to the New Relic Intelligent Observability Platform promise to give DevOps and platform engineering teams the context needed to more proactively IT environments using AI agents that will also be integrated with platforms such as Google Gemini and ServiceNow.
At the same time, New Relic is also adding a Transaction 360 tool that makes use of traces to observe specific business events, along with an Engagement Intelligence tool to track user behavior.
Finally, New Relic is also adding digital experience management (DEM) tools to track, for example, video metrics across a diverse range of devices and accessing backend services located in multiple regions.
Nic Benders, chief technical strategist for New Relic, said the goal is to apply multiple classes of predictive and generative AI models to telemetry data in a way that makes it easier for DevOps and platform engineering teams to surface actionable insights to, for example, reduce cloud costs, across application environments that have become too complex to otherwise manage.
Pulling telemetry data into observability isn’t enough, however, because IT professionals are not going to know how to query that data in a way that surfaces meaningful insights, said Benders. IT teams need to be augmented by AI models that enable them to both prevent issues from arising in the first place while adroitly responding to incidents before they broadly affect the application environment, noted Benders. In effect, DevOps and platform engineering teams are moving into a new era of intelligence, he added.
New Relic has been investing in a broad range of AI technologies that it has previously made available to more than 85,000 customers. The New Relic platform is providing the core foundation for orchestrating those capabilities in a way that ultimately makes it simpler to rely on multiple AI agents and models to automate various tasks, from either directly in an existing workflow or the user interface provided by the New Relic platform, said Benders.
There is, of course, already no shortage of options for applying AI to DevOps workflows. The level of sophistication of those AI tools will naturally vary widely but the fact remains that more tasks will soon be automated using AI agents and models. Each DevOps and platform engineering team will need to determine which tasks might lend themselves better to being allocated to an AI agent versus continuing to be performed by a software engineer.
Ultimately, the goal remains to safely deploy as much software as quickly as possible. It’s not likely that will be achieved any time soon relying solely on AI, however, as software engineering continues to evolve the current level of toil currently being experienced by application development teams should continue to steadily decline. The issue now is making sure expectations actually align with what AI agents based on probabilistic AI models are actually capable of reliably doing within the context of a DevOps workflow.
