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

# Introduction

> Welcome to Deeptrace, your 24/7 AI SRE and debugging co-pilot that automatically investigates and root-causes production alerts by reasoning across code, observability, and infrastructure. 

Deeptrace operates like your team's senior engineers — correlating logs, traces, metrics, and code to tell your engineers what’s actually happening in production. Engineering teams use Deeptrace to reduce MTTR, save thousands of engineering hours annually, and stay ahead of customer-facing issues.

## How Deeptrace works

<Steps>
  <Step title="Understands Production">
    Once integrated, Deeptrace builds an internal mapping across your observability data and codebase to understand how your production system truly behaves and uncover hidden assumptions.
  </Step>

  <Step title="Investigates alerts automatically">
    Deeptrace launches an investigation on every alert that hits your engineering Slack channel. Within minutes, Deeptrace provides a clear root cause summary along with ground-truth sources to validate its findings.
  </Step>

  <Step title="Takes action">
    Deeptrace turns its findings into actionable steps. You can chat with Deeptrace to further debug an issue, create a PR to ship a fix, open a ticket that centralizes all relevant context, or configure rules to reduce alert noise.
  </Step>

  <Step title="Learns over time">
    Deeptrace refines its understanding using your feedback, prior investigations, and historical context, so future alerts are resolved faster and with greater accuracy.
  </Step>
</Steps>

## Impact with Deeptrace

Deeptrace is trusted by fast-growing startups and Fortune 1000 companies, helping engineering teams resolve incidents faster, reduce on-call toil, and operate reliably at scale.

<Tabs>
  <Tab title="MTTR">
    Across teams, Deeptrace cuts mean time to resolution **(MTTR) by \~50%** by reducing triage time and eliminating manual context gathering for every alert.
  </Tab>

  <Tab title="Engineering Hours">
    Across teams, Deeptrace saves **thousands of engineering hours** annually by reducing time spent on debugging, triaging, and maintaining production, allowing engineers to focus on higher-leverage work.
  </Tab>

  <Tab title="Investigations">
    Across teams, Deeptrace investigates **25K+ production alerts** every month, becoming a core part of on-call and debugging workflows.
  </Tab>

  <Tab title="Accuracy">
    Across teams, Deeptrace delivers **\~70% accuracy** across investigations, providing reliable root causes within minutes.
  </Tab>
</Tabs>

## Dive in!

<Columns cols={2}>
  <Card title="Get Started" icon="rocket" href="/integrations">
    Create an account, connect your tools, and start using Deeptrace in minutes.
  </Card>

  <Card title="Guides" icon="book" href="/guide">
    Learn how to investigate alerts and collaborate with Deeptrace in Slack.
  </Card>

  <Card title="Features" icon="sparkles" href="/features">
    Explore Deeptrace’s features, from rule-based filtering to custom context.
  </Card>
</Columns>