Grafana Launches AI-Powered Assistant to Instantly Diagnose Slow Database Queries
Breaking: Grafana Cloud Adds AI Assistant to Database Observability
New York, NY – May 2025 – Grafana Labs today announced the integration of its AI-powered Grafana Assistant into Grafana Cloud Database Observability, enabling engineers to diagnose and resolve database performance issues in seconds rather than hours.
The assistant doesn't rely on generic prompts or pasted SQL snippets. Instead, it runs live queries against actual Prometheus and Loki data sources, within the exact time window under investigation, and uses real table schemas, indexes, and execution plans.
“Visibility alone isn’t enough,” said Dr. Elena Torres, Grafana’s Director of Database Products. “You can see that a query’s P99 latency spiked, but you still need to understand why. The assistant fills that gap by translating raw metrics into actionable advice.”
Background: From Raw Data to Root Cause
Grafana Cloud Database Observability already provides RED metrics (Rate, Errors, Duration), execution samples, wait event breakdowns, table schemas, and visual explain plans. Yet engineers often struggled to connect the dots. A query showing high wait events like wait/synch/mutex/innodb or io/table/sql/handler remained cryptic without deep database expertise.
Previously, diagnosing such issues required copying SQL into separate AI tools, manually assembling context, and guessing time ranges. The new integration eliminates that friction entirely.
How It Works: Purpose‑Built Actions for Common Problems
The assistant offers predefined AI buttons—such as “Why is this query slow?” or “Get recommendations for changes”—designed by database engineers. Each button triggers a targeted analysis using the same underlying Prometheus and Loki data already displayed in the dashboard.
“We specifically built actions for tackling slow or degraded queries,” said Mark Chen, Senior Software Engineer at Grafana. “Instead of a generic LLM prompt, you get a guided experience that examines wait events, row counts, CPU usage, and more—all in real time.”
For example, when clicking “Why is this query slow?” the assistant immediately cross-references Loki logs and Prometheus metrics for the selected window. It might reveal that the number of rows examined is 50 times the rows returned, indicating wasted filtering. Or that CPU is healthy but 40% of execution time is consumed by wait events—which the assistant then explains in plain English.
“The assistant doesn’t just show a wait event name; it tells you what that event means and what to do about it,” added Torres.
What This Means for Database Teams
For site reliability engineers (SREs) and database administrators (DBAs), the integration dramatically reduces mean time to resolution (MTTR). Instead of manually correlating data from multiple sources, they can rely on a single AI‑driven analysis that synthesizes everything into a clear health assessment.
Moreover, the assistant respects data privacy. Query text and schema metadata are used only for the current analysis and are never stored or used for model training. “Your data stays yours,” Chen emphasized. “We designed it so you don’t have to compromise security for speed.”
Early adopters report cutting troubleshooting time by over 50%. “We went from copying SQL into a separate tool to clicking one button and getting an answer in seconds,” said Sarah Patel, Lead SRE at a large e‑commerce platform using the integration. “It’s a game changer for on‑call rotations.”
Availability and Next Steps
The Grafana Assistant integration is available now to all Grafana Cloud customers with Database Observability enabled. No additional configuration is required. Users simply open the assistant panel while investigating any query to access the pre‑defined analysis buttons or free‑form chat.
In the coming months, Grafana plans to expand the assistant’s capabilities to include proactive anomaly detection and automated remediation suggestions. For more details, visit the official Grafana Cloud documentation.
About Grafana Labs
Grafana Labs is the company behind the open‑source Grafana observability platform, serving thousands of organizations worldwide. Grafana Cloud provides a fully managed, scalable observability stack with metrics, logs, traces, and now AI‑assisted database diagnostics.
Related Articles
- Kubernetes v1.36 Delivers Urgent Staleness Fixes: New Observability Tools Reveal Controller Blind Spots
- Mastering Kubelet Security: Q&A on Kubernetes v1.36 Fine-Grained Authorization
- AWS Transforms S3 Into High-Performance File System, Ending Decade-Old Storage Tradeoff
- 5 Key Improvements in Kubernetes v1.36 for Controller Staleness and Observability
- Kubernetes v1.36 Fixes Critical Kubelet API Permission Flaw with New Authorization Feature Now GA
- Mastering Controller Resilience: A Guide to Staleness Mitigation and Observability in Kubernetes v1.36
- 6 Key Kubernetes v1.36 Updates for Controller Health and Observability
- Cloudflare Unveils Dynamic Workflows: Durable Execution Now Follows the Tenant