New Relic is unveiling an observability solution aimed at applications that run in ChatGPT, Techzine reported on Jan. 23.

New Relic aims to solve performance and user-behaviour “black box” problems that arise in generative AI environments.

The solution is designed to help track how apps actually work and how they interact with users, as more enterprise applications use AI as an interface.

Companies are providing users with more features through ChatGPT, but the company said this is making it harder for existing browser-based monitoring tools to detect performance issues, layout errors, button malfunctions and when users drop off.

It added that, given the nature of GPT environments, AI-generated interfaces may look fine but often produce errors or show data not connected to the backend.

New Relic expanded its browser agent so it can collect PageViews, AjaxRequests and console errors even in iFrame environments inside ChatGPT.

It said this allows real-time visibility into user latency, connection errors and script errors caused by AI output, and that development teams can use the data to define key interactions and set tracking points themselves.

It said interactions that occur in ChatGPT apps can also be connected through to backend systems to secure end-to-end visibility.

The feature is included in New Relic’s existing Intelligent Observability Platform and is available for immediate use.