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

# Getting Started

> Everything you need to integrate the WorkBuddy Voice API into your application.

## Prerequisites

Before you begin, make sure you have:

* A WorkBuddy kiosk API key (provided by your account admin)
* The API base URL for your environment
* Familiarity with WebSocket connections and audio streaming concepts

## For Web Applications

No additional packages are required. The integration uses native browser APIs:

| API           | Purpose                               |
| ------------- | ------------------------------------- |
| WebSocket     | Real-time bidirectional communication |
| Web Audio API | Microphone capture and audio playback |
| getUserMedia  | Access to the device microphone       |

Supported browsers: Chrome, Edge, Firefox, Safari (latest versions).

## For Flutter Applications

Add these dependencies to your `pubspec.yaml`:

```yaml theme={null}
dependencies:
  web_socket_channel: ^2.4.0   # WebSocket client
  record: ^5.0.0               # Microphone capture (PCM)
  flutter_sound: ^9.0.0        # Audio playback (PCM stream)
  http: ^1.0.0                 # HTTP client for token endpoint
  permission_handler: ^11.0.0  # Runtime permission requests
```

## Security Best Practices

Your API key grants access to voice sessions on your account. Treat it as a sensitive credential.

* Never hardcode API keys in client-side source code
* Use environment variables or secure configuration management
* For web apps, proxy the token request through your own backend to avoid exposing the key
* For mobile/desktop apps, use encrypted storage or environment files
* Add `.env` to your `.gitignore`

## Integration Overview

The integration follows a simple three-step flow:

1. **Authenticate** — POST your API key to get a short-lived session token
2. **Connect** — Open a WebSocket and send the token as the first message
3. **Stream** — Send microphone audio and receive AI responses as raw PCM frames

Continue to the [Authentication](/authentication) page to learn how to obtain your session token.