Grafana Labs unveils AI agents to automate AI observability

The company introduced Assistant Watchers, always-on agents that observe and learn the normal telemetry patterns of other AI agents and flag anomalies before larger problems emerge. It also launched Grafana Cloud Agents in preview, which provide a controlled environment for AI to analyze code in the context of its production behavior and prepare software fixes that humans can review. A new capability called Pages lets teams build customized, app-like visualizations of AI workloads that go beyond standard dashboards.

Grafana said the products support a shift toward "agentic operations." Mat Ryer, Grafana's senior director of AI, said enterprises want AI agents to take over more work, but need evidence and control at each step. "Watchers surface emerging problems, investigations help diagnose them, and Cloud Agents prepare fixes for human review," Ryer said in the announcement. "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 approach treats agent sessions and large language model conversations as standard telemetry signals, similar to logs, metrics and traces from traditional applications. The company said its original stack was built for cloud infrastructure and older software, while generative AI applications behave differently and have been difficult for legacy observability tools to debug.

Beyond AI monitoring, Grafana introduced query performance enhancements in Grafana Cloud Logs based on technology from its recent acquisition of Logline. The company said the update enables faster full-text search across low-frequency data, such as specific request IDs, at no additional cost. It also announced Adaptive Logs Archive within its Adaptive Telemetry Suite, which lets enterprise users route logs to their own object storage buckets using open data formats to retain observability data at lower cost for future investigations and compliance checks.

The company also announced Knowledge Graph and Digital Experience Monitoring capabilities intended to connect infrastructure and application dependencies with real user experiences, giving AI agents and human engineers a shared runtime context during investigations. New Mobile Observability and Session Replay tools are meant to help teams see issues from the user's perspective.

Richard Lam, Grafana's director of product, said most teams do not have a monitoring problem but a technology debt problem: telemetry from every layer with no way to connect it to the affected customer. He said putting a knowledge graph in every Grafana Cloud stack and extending it from infrastructure to the user makes unified observability foundational instead of something each team must assemble on its own.

Editor's Summary

Grafana Labs introduced Assistant Watchers, Grafana Cloud Agents and other updates at ObservabilityCON 2026, aiming to automate AI workload monitoring and remediation. The company also added lower-cost log search and storage options plus knowledge-graph and digital-experience tools. The announcements show Grafana pushing to unify AI and traditional infrastructure observability in one platform.