Overview
Not using one of our pre-built integrations? No problem. Roark’s API and SDKs make it simple to send call data from any voice AI platform or custom application. At its core, you just need to send us an audio file and we’ll handle the rest.Integration Options
Direct API
Send call data directly to our REST API endpoints:Node.js SDK
Install and use our Node SDK for a more integrated experience:Python SDK
Use our Python SDK for seamless integration:Core Capabilities
With custom integrations, you can:Send Call Data
- Audio files - MP3, WAV, or any common audio format
- Transcripts - Pre-transcribed text if available
- Metadata - Agent info, customer details, call context
- Tool calls - Track function invocations and API calls
Run Evaluations
- Apply built-in evaluators
- Create custom evaluation criteria
- Get scores and insights
- Track performance over time
Manage Agents
- Register custom agents
- Track agent versions
- Compare performance
- Use agents in simulations
Getting Started
1
Generate API Key
Create an API key in your Roark dashboard
2
Choose Integration Method
Select direct API, Node.js SDK, or Python SDK
3
Send Your First Call
Upload an audio file and metadata using your chosen method
4
View Results
Check your Roark dashboard for analysis and insights
Audio Requirements
For best results, ensure your audio files meet these specifications:| Requirement | Specification |
|---|---|
| Format | MP3, WAV, M4A, or FLAC |
| Sample Rate | 8kHz or higher |
| Channels | Mono or stereo |
| Duration | Up to 2 hours |
| File Size | Max 500MB |
| Access | Publicly accessible URL or base64 encoded |
Need a Custom Integration?
If you need deeper integration with a specific platform or have unique requirements:Contact Our Team
Reach out to discuss custom integration options
- Native integrations for unsupported platforms
- Custom webhook handlers
- Bulk migration from other systems
- Enterprise-specific requirements