Grafana Labs leans on AI agents to automate AI agent observability

Grafana Labs leans on AI agents to automate AI agent observability

Observability software pioneer Grafana Labs Inc. is flexing its muscles in the rapidly emerging market for artificial intelligence monitoring tools with a swath of updates to its Grafana Cloud platform announced at ObservabilityCON 2026 today.

The startup announced a new suite of AI agents, including Assistant Watchers for monitoring AI workloads and autonomous agents, and Grafana Cloud Agents for remediating problems, in addition to various broader enhancements focused on monitoring and understanding AI telemetry data.

Grafana first made a name for itself in traditional observability with its open-source Grafana platform, and since the turn of the year has aggressively pivoted to focus specifically on AI observability. The original Grafana stack was built to monitor cloud infrastructure and the software applications of yesteryear, and the startup was one of the first to recognize that generative AI applications behave very differently.

Because traditional apps work so differently, legacy observability tools struggled to provide much insight into AI models and agents, making them difficult to debug. Grafana’s solution is to treat agent sessions and large language model conversations as standard telemetry signals, similar to the logs, metrics and traces associated with traditional applications. That way, customers can monitor the performance of these systems in the context of their broader information technology infrastructure.

With the rise of autonomous agents that perform increasingly complex tasks on behalf of humans, Grafana is responding with the launch of highly specialized AI agents that are specifically designed to observe the behavior of those systems. The company said it’s pushing towards “agentic operations,” and a key part of this push is its new Assistant Watchers, which are always-on agents designed to observe and learn the normal telemetry patterns of other agents and flag anomalies in their behavior before bigger problems arise.

Alongside the Assistant Watchers, it’s also launching Grafana Cloud Agents in preview, which provide a controlled environment where AI can analyze code in the context of its production behavior and prepare software fixes that can be reviewed by humans. Meanwhile, Pages is a new capability that enables teams to build customized, app-like visualizations of AI workloads that provide deeper insights than its standard dashboards.

Grafana Senior Director of AI Mat Ryer said enterprises want AI agents to take over more work from humans, but if they’re going to do this, they need both evidence and control at each step. “Watchers surface emerging problems, investigations help diagnose them, and Cloud Agents prepare fixes for human review,” he said. “Together they make Grafana Cloud more autonomous, bringing detection, diagnosis and proposed action into one workflow, grounded in the team’s own telemetry.”

Grafana’s agents might be quite sophisticated, but the company recognizes that observability needs to do more than just secure AI workloads. One problem it’s seeking to address is the cost of running AI in the cloud. To do this it’s introducing query performance enhancements in Grafana Cloud Logs, based on technology it obtained through its recent acquisition of Logline. This enables faster full-text search across low-frequency data such as specific request IDs at no additional cost, Grafana said.

Another update is the new Adaptive Logs Archive within Grafana’s Adaptive Telemetry Suite, which allows enterprise users to route logs to their own object storage buckets using open data formats so they can retain observability data at lower cost for future investigations and compliance checks.

Finally, Grafana said, it’s trying to address full-stack visibility across both AI and traditional cloud workloads with the introduction of its new Knowledge Graph and Digital Experience Monitoring capabilities. Knowledge Graph can connect infrastructure and application dependencies with real user experiences, so that both AI agents and human engineers can view a shared runtime context when investigating problems. This capability is augmented by new Mobile Observability and Session Replay tools that enable teams to see issues from the user’s perspective.

“Most teams do not have a monitoring problem,” said Director of Product Richard Lam. “What they have is a technology debt problem: telemetry from every layer and no way to connect it to the customer it affects. Putting a knowledge graph in every Grafana Cloud stack and extending it from the infrastructure to the user makes unified observability foundational instead of something every team has to assemble on its own.”

Today’s announcements indicate that Grafana has a clear vision for what is fast becoming confusing future, in which AI is used for observability and observability is a must for AI.

Image: SiliconANGLE/Microsoft Designer