---
title: "Connective intelligence layer | Grafana Labs"
description: "Automatic issue detection and guided troubleshooting across your entire system."
---

> For a curated documentation index, see [llms.txt](/llms.txt). For the complete documentation index, see [llms-full.txt](/llms-full.txt).

## Connective intelligence with Knowledge Graph

Knowledge Graph is the connective intelligence foundation built into Grafana Cloud:

- The engine that powers the opinionated observability experience across turnkey solutions (Kubernetes Monitoring, Application Observability, Database Observability, and Frontend Observability)
- Automatically discovers your services, infrastructure, and databases, maps how they depend on each other
- Continuously analyzes your environment for issues
  
  [Knowledge Graph: automatic mapping, detection, and investigation](knowledge-graph-overview.svg "Knowledge Graph: automatic mapping, detection, and investigation")

[Try it out in Grafana Play](https://play.grafana.org/a/grafana-asserts-app/catalog?start=now-1h&end=now&searchQuery=&entityTypeFilter=Service&pageSize=15&page=1&sortBy%5Bid%5D=AssertionCircle&sortBy%5Border%5D=desc&columns%5B0%5D%5Bid%5D=Cpu&columns%5B1%5D%5Bid%5D=Memory&columns%5B2%5D%5Bid%5D=Disk&uninstrumentedFilter=instrumented&otelServiceFilter=all)

## Core capabilities

- [**Entity Catalog**](/docs/grafana-cloud/knowledge-graph/troubleshoot-infra-apps/explore-entity-catalog/), where you can view all entities (services, hosts, pods, databases) and their relationships
- [**RCA (root cause analysis) Workbench**](/docs/grafana-cloud/knowledge-graph/troubleshoot-infra-apps/workbench/), where you view correlated metrics, logs, traces, and insights on a timeline for guided investigation
- [**Insights**](/docs/grafana-cloud/knowledge-graph/reference/insights-categories/), which surfaces five categories of issues:
  
  - **Saturation**: Resources approaching limits (CPU, memory, disk, connections)
  - **Amend**: Deployments, config changes, scale events
  - **Anomaly**: Unusual patterns in traffic or behavior (request rate spikes or drops, latency changes)
  - **Failure**: System deviating from desired state (pod crashes, missing Redis master, controller errors)
  - **Error**: Problematic requests (5xx errors, latency threshold breaches, error log rate breaches)