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

# Introduction

> AI meeting copilot with real-time transcription, echo cancellation, and a stealth overlay invisible to screen sharing.

<img className="block dark:hidden" src="https://mintcdn.com/laxcorpresearch-3b85a1df/LgQt-X7yep1rjG92/images/architecture.png?fit=max&auto=format&n=LgQt-X7yep1rjG92&q=85&s=b25799623390a90b5cbf4dc9552240da" alt="Raven Architecture" width="7037" height="3738" data-path="images/architecture.png" />

Raven is an AI-powered meeting copilot that runs on your desktop. It captures both sides of any conversation, cancels echo so voices don't bleed between streams, transcribes everything in real-time, and gives you AI assistance with full context - all while staying completely invisible to screen sharing.

## Get Raven

<CardGroup cols={2}>
  <Card title="Download Raven" icon="download" href="https://useraven.ai">
    Download the app for macOS or Windows. Sign in and start using Raven immediately - no setup required.
  </Card>

  <Card title="Build from Source" icon="code" href="/quickstart/overview">
    Clone the open-source repo, bring your own API keys, and build Raven yourself.
  </Card>
</CardGroup>

## Features

<CardGroup cols={2}>
  <Card title="Dual-Stream Audio" icon="waveform-lines">
    Captures system audio and microphone simultaneously on macOS and Windows using native platform APIs.
  </Card>

  <Card title="Echo Cancellation" icon="wave-pulse">
    Built-in echo cancellation prevents the remote speaker's voice from leaking into your microphone stream.
  </Card>

  <Card title="Real-Time Transcription" icon="closed-captioning">
    Live speech-to-text with separate speaker labels for "You" and "Them". Supports 20+ languages.
  </Card>

  <Card title="AI Assistance" icon="brain">
    Get context-aware help from Claude or GPT during meetings - quick actions, follow-ups, recaps, and more.
  </Card>

  <Card title="Screenshot Context" icon="camera">
    Press `Cmd+Enter` to capture your screen and send it to the AI. Solve problems, interpret slides, or analyze anything visible.
  </Card>

  <Card title="Stealth Overlay" icon="eye-slash">
    Completely invisible to Zoom, Meet, Teams, and Discord screen sharing. Drag, resize, and interact while staying hidden.
  </Card>

  <Card title="Local-First" icon="lock">
    Your data stays on your machine. Optional incognito mode for sessions you don't want saved.
  </Card>

  <Card title="Modes" icon="sliders">
    Customizable AI profiles — Interview, Sales, Meeting Notes, Job Search, Learning, Coding Interview — plus document upload (RAG) for extra context.
  </Card>

  <Card title="Session History" icon="magnifying-glass">
    Every session is saved with AI-generated titles and summaries. Search across all past meetings by keyword.
  </Card>

  <Card title="20+ Languages" icon="language">
    Transcription in English, Hindi, Spanish, French, German, Japanese, Korean, Chinese, Arabic, and more. Output language configurable independently.
  </Card>

  <Card title="Meeting Detection (Pro)" icon="video">
    Auto-detects Zoom, Teams, and Meet meetings on Apple Silicon Macs. Post-meeting insights including summaries, action items, and sentiment analysis.
  </Card>

  <Card title="Cloud Sync (Pro)" icon="cloud">
    Sessions, modes, and documents sync across devices. Sign in with Google or email.
  </Card>
</CardGroup>

## How It Works

<Steps>
  <Step title="Start recording">
    Click the record button on the overlay or press `Cmd+R`.
  </Step>

  <Step title="Audio capture">
    Raven captures both system audio (what you hear) and your microphone simultaneously using native platform APIs.
  </Step>

  <Step title="Echo cancellation">
    Both streams pass through an echo cancellation pipeline so the remote speaker's voice doesn't contaminate your mic signal.
  </Step>

  <Step title="Transcription">
    Clean audio is transcribed in real-time. Your speech appears as "You" and the remote speaker as "Them".
  </Step>

  <Step title="AI assistance">
    Ask the AI for help at any time - it has full context from the conversation. Press `Cmd+Enter` to include a screenshot of your screen for visual context.
  </Step>
</Steps>

## Platform Support

| Platform                              | Status            |
| ------------------------------------- | ----------------- |
| **macOS 12+** (Intel & Apple Silicon) | Fully supported   |
| **Windows 10/11**                     | Supported         |
| Linux                                 | Not yet supported |

## Keyboard Shortcuts

| Action                      | Shortcut                |
| --------------------------- | ----------------------- |
| Toggle Overlay              | `Cmd + \`               |
| Start/Stop Recording        | `Cmd + R`               |
| AI Assist (with screenshot) | `Cmd + Enter`           |
| Clear Conversation          | `Cmd + Shift + R`       |
| Move Overlay                | `Cmd + Arrow Keys`      |
| Scroll Overlay              | `Cmd + Shift + Up/Down` |

<Note>On Windows, replace `Cmd` with `Ctrl`.</Note>

## Next Steps

<CardGroup cols={2}>
  <Card title="Using Raven" icon="play" href="/features/recording">
    Learn how recording, transcription, and AI assistance work.
  </Card>

  <Card title="Raven Pro vs Open Source" icon="crown" href="/pro/free-vs-pro">
    Compare the Pro app with the open-source version.
  </Card>

  <Card title="Build from Source" icon="terminal" href="/quickstart/overview">
    Set up the open-source version on macOS or Windows.
  </Card>

  <Card title="Contributing" icon="code" href="/contributing/repo-structure">
    Understand the repo structure and how to contribute.
  </Card>
</CardGroup>