Operations | Monitoring | ITSM | DevOps | Cloud

Some questions in engineering never go away. Here’s one that every team eventually confronts: Do we roll up our sleeves and build the tooling ourselves, or do we buy something built for us? It’s a choice that has the power to speed teams up or hold them back. With the rise of AI-powered observability, this familiar software dilemma has re-emerged with higher stakes and faster-moving technology.

Continuous profiling offers deep visibility into production environments, revealing exactly how applications consume CPU and memory. It’s the go-to observability practice for directly connecting system behavior and performance to specific lines of code. But when teams consider deploying continuous profiling more broadly, a common question comes up: what’s the overhead? Is it safe to run continuous profiling on my production services 24/7, or does the cost outweigh the benefits?

Note: This blog post originally published in May 2025 and was updated in February 2026 to reflect that Git Sync is now available in public preview in Grafana Cloud. As your Grafana instance scales, so does the challenge of maintaining dashboards. Managing dozens—or hundreds—of dashboards through the UI alone can quickly become overwhelming. Tracking changes gets murky, dashboards multiply, and consistency suffers.

Grafana Assistant is the most general-purpose tool we’ve delivered since dashboards. People use our Grafana Cloud LLM to understand unfamiliar areas of their stacks, generate dashboards and beautiful visualizations out of thin air, build queries, and support investigations.

As agentic AI applications are used more broadly in production, they introduce new operational models, combining multi-step reasoning, tool execution, and autonomous decision-making into a single workflow. SRE teams need visibility into how these agents behave, where they fail, and how they perform over time.

Let’s be real: AI can hallucinate. And in observability, that feels risky. No one wants an assistant that sends your SREs chasing ghosts. At best, that burns expensive engineering time. At worst, it slows incident response in production and pushes teams toward the wrong remediation path. So here’s the big question: What makes Grafana Assistant different, and why should you trust it? Let’s start by acknowledging the fear. AI hallucinations are a real issue.

Back in 2021, Grafana Labs CTO Tom Wilkie (then VP of Products) spoke at PromCON about the need for improvements in Prometheus' remote write capabilities. “We use between 10 and 2 bytes per sample to send via remote write, and Prometheus only uses 1 or 2 bytes per sample on the local disk so there’s big, big room for improvement,” Wilkie said at the time.

Sometimes the most interesting engineering stories don’t start with a roadmap or a release plan—they start with personal taste. A preference for good design. A frustration with clunky tools. A desire to see everything in one place.

Tempo 2.10 has arrived, delivering TraceQL enhancements, improved cardinality management for the metrics-generator, vParquet5, and more. You can continue reading and check out the video below to learn more about these and other new features. The Tempo 2.10 release notes and changelog provide more in-depth details and include all of the changes that came with this release.

Volkov Labs has been a longtime partner to Grafana Labs, with co-founder Mikhail Volkov contributing to Grafana in the early stages of the OSS project. On Sept. 26, the Florida-based company that recently created a suite of business intelligence (BI) plugins for Grafana announced it had been acquired. In light of the news, Grafana Labs committed to taking over the maintenance and development of their popular business intelligence (BI) plugin suite.