> ## Documentation Index
> Fetch the complete documentation index at: https://docs-sre.lightrun.com/llms.txt
> Use this file to discover all available pages before exploring further.

# AI SRE Overview

> What AI SRE does and how it works

AI SRE is an AI-powered incident response assistant that helps support engineers, production engineers, on-call engineers, application engineers, and SREs investigate and resolve production incidents faster.

## Overview

AI SRE investigates incidents by:

* **Accessing code** — Reviews code changes, commits, and correlates them with incidents
* **Querying systems** — Gathers evidence from telemetry, infrastructure, and knowledge bases
* **Building evidence chains** — Correlates findings across systems to identify root causes
* **Delivering insights** — Provides evidence-based conclusions with confidence levels and suggested actions

## Benefits

* **Faster MTTR** — AI SRE does the investigation work, reducing investigation time from hours to minutes
* **Evidence-based** — Uses facts from your systems, not speculation—every finding is backed by data from your code, logs, and metrics
* **Works with your tools** — No data migration required; connects to your existing stack
* **Confidence levels** — Clearly states certainty and what's missing, so you know when to verify

## Architecture

```mermaid theme={null}
graph LR
    A[You ask question] --> B[AI SRE gathers evidence]
    B --> C[Correlates data]
    C --> D[Provides findings]
    D --> E[You take action]
```

1. **You ask** — Type questions about incidents in natural language
2. **AI SRE works** — Queries code repositories, scans logs and metrics, reviews recent changes, and correlates data across your stack
3. **AI SRE correlates** — Links findings across systems, builds evidence chains, and identifies root causes
4. **AI SRE delivers** — Provides evidence-based conclusions with confidence levels and actionable next steps
5. **You resolve** — Use insights to fix incidents faster

## Use cases

* **"Explain the latest code change and its impact"** — AI SRE reviews recent deployments and code changes, correlating with current system state to explain what changed and how it affects your services

* **"Analyze this alert and summarize what's happening"** — AI SRE queries logs, metrics, and traces to understand the alert context, identifies affected services, and provides a clear summary of what's broken

* **"What caused this incident?"** — AI SRE investigates by reviewing code changes, correlating with incident timeline, querying telemetry data, and building evidence chains to identify the root cause

* **"Which team owns this service?"** — AI SRE searches code repositories, reviews ownership patterns, and identifies team information from your knowledge base

## Data Security & Privacy

Your data is isolated per company, encrypted in transit and at rest, and never used for AI model training. Integrations use read-only access. See the [Security & Privacy](/faq#security--privacy) section for details.

## Limitations

* **Read-only** — AI SRE cannot make changes to your systems
* **Data availability** — Insights depend on connected integrations and data quality
* **Confidence levels** — Findings include confidence levels; always verify critical decisions
* **Tool expertise** — Works best with well-configured integrations and monitoring

## Get started

<CardGroup cols={2}>
  <Card title="Get Started" icon="rocket" href="/getting-started/overview">
    Set up AI SRE
  </Card>

  <Card title="Workflows" icon="wrench" href="/working-with-ai-sre/overview/detection">
    Learn workflows
  </Card>
</CardGroup>