Operations | Monitoring | ITSM | DevOps | Cloud

Actually useful AI is welcome in observability. AI for the sake of AI is not. In this overview of Grafana Labs’ 4th annual Observability Survey, Marc Chipouras shares what 1,300+ respondents from 76 countries told us about the current state of observability — and what comes next. This year’s survey explores four major themes: The results show strong interest in AI for forecasting, root cause analysis, onboarding, and generating dashboards, alerts, and queries. But when it comes to autonomous action, practitioners are more cautious — and 95% say AI needs to show its work to earn trust.

In this episode, Mat Ryer and Cyril Tovena will talk about agentic LLMs with host Tiffany Jernigan.

From "something's off" to "here's why (and how to fix it)." Grafana Assistant goes full detective mode: pulls the clues, connects the dots, and recommend a fix... with receipts.

In this video, Collin Fingar, Software Engineer at Grafana Labs, demonstrates how template variables can be used in saved queries, a feature that enables users to reuse queries they or others in their org have saved. You'll see how a query that contains variables can be reused, and how the variables can be replaced at the point of reuse.

As part of the Grafana Everywhere series, Chris Hodges from Dropbox shares why they chose Loki for their logging stack after their data center went dark.

You add the OpenTelemetry Java agent, restart your app - and like magic, observability appears. But is it really magic? What’s actually enabled by default? What telemetry should you expect to see? What’s missing? And what might you want to tweak, tune, or even turn off?

At today’s community call, we will hear from David Grant, one of the engineers who has brought a new component, the Block Builder, into Mimir. Using the Ingest Storage architecture in Mimir 3.0, the Block Builder takes over the block-building responsibility from the Ingester. This feature is experimental in Mimir today, but is rolling out to production inside of Grafana Labs now. This is a great time to introduce the component, discuss the motivation, and show where it fits in the larger architecture.

Grafana Campfire Community Calls are back. We are starting with *Back to the Basics.* Even though you heard it so many times, but some of you are either new to or not very experienced with terms such as monitoring, observability, metrics, tracing, profiling etc. The good news is that you're not alone, and we've got this!! This will be a perfect learning opportunity to gain understanding during this live call and ask questions if anything is not clear.

In the February edition of the Kubernetes Monitoring Helm chart office hours, we discuss the version 3.8 release, the upcoming 4.0 release and features, and we discuss the upcoming deprecation of the 1.x and 2.0 versions.

SREs are famously skeptical of AI — so how do you convince them to trust agents in production? In this episode of Grafana’s Big Tent, Tom Wilkie talks with Spiros Xanthos (Resolve AI), Manoj Acharya (Grafana Labs), and Cyril Tovena (Grafana Assistant team) about agent-first observability. They unpack knowledge graphs, LLM reasoning, autonomous debugging, pricing models, and the “Claude Code moment” for observability. Is autonomous production ops closer than we think?