Run data analysis and investigations

Use this hub when you need to understand live system behavior or explain an emerging incident. The workflows are self-contained so you can start wherever you are, whether to ask a quick question, launch an investigation, correlate multiple signals, or refresh infrastructure context, without reading prerequisite chapters.

When to use this hub

You want fast answers about metrics, logs, traces, profiles, or SQL data without leaving Grafana.
An incident requires deeper, parallel analysis using Assistant investigations.
You need the Assistant to incorporate infrastructure topology before suggesting next steps.

Prerequisites

Connected data sources: Configure Prometheus, Loki, Tempo, Pyroscope, SQL, or other data sources in Grafana Cloud and keep them actively collecting data.
Assistant access: Enable Grafana Assistant for your organization and confirm you have permission to open the Assistant sidebar.
Prompting etiquette: Use natural language, be specific about timeframes and services, provide examples when possible, and start a new conversation when the topic changes.

Workflows in this hub

Answer quick questions with chat
Use Grafana Assistant chat to query observability data and explain findings.
Correlate multiple signals
Combine metrics, logs, traces, profiles, and SQL insights in a single investigation.
Run a systematic investigation
Launch and guide multi-agent workflows with Grafana Assistant Investigations.
Use infrastructure memory for context
Refresh, review, and apply infrastructure memories to investigations and chat workflows.