Overview
Organizations with strict compliance requirements such as those needing to keep voice audio content within a specific region often ask how Twilio handles call audio when using Media Streams and how to ensure no audio is stored outside their country. This article explains how to architect your Twilio solution to keep audio content transient and regionally processed, specifically using the JP1 (Tokyo) Edge, and how to disable Voice Trace for additional compliance.
Key Points
- Twilio Media Streams allow you to stream real-time audio from voice calls to your own WebSocket server.
- By using the JP1 Edge (Tokyo), you can ensure media processing occurs locally in Japan before audio is streamed to your infrastructure.
-
No audio files are stored by Twilio if you do not use the
<Record>verb in your TwiML. - Voice Trace is an optional troubleshooting feature that, if enabled, may temporarily store audio for diagnostic purposes. Disabling it is recommended for compliance.
Product
Programmable Voice
Environment
Twilio Console
Procedure
1. Streaming Audio Without Storage
- Use Twilio Media Streams to send call audio to your own WebSocket endpoint.
-
Do not use the
<Record>verb in your TwiML. This ensures Twilio does not create or store any audio recordings. - Only call metadata (such as logs and billing information) is handled in the US1 region; audio content remains transient.
2. Using the JP1 Edge for Local Media Processing
- Configure your application to use the Tokyo Edge (JP1).
- This ensures that media processing (the gateway handling the audio packets) occurs at the Tokyo edge before streaming to your WebSocket server in Japan.
- For inbound calls, set up regional number routing to anchor calls in JP1.
- For outbound calls, use regional APIs and credentials to originate calls from JP1.
3. Disabling Voice Trace
To further ensure compliance, disable the Voice Trace feature in your Twilio Console:
- Log in to your Twilio project at Twilio Console. If you are using a subaccount, select it as needed.
- Navigate to the Voice Settings page.
- Scroll down to the Voice Trace section and select Disabled.
- Click Save to apply the changes.
For more details, see: Enable Voice Trace for Troubleshooting Twilio Call Audio and DTMF Issues
Compliance Summary
By following the above steps:
- No audio content is stored by Twilio, all audio is streamed in real time and remains transient.
- Media processing occurs in Japan (when using JP1 Edge).
- Voice Trace is disabled, preventing any diagnostic audio storage.
This architecture helps organizations meet strict regional compliance requirements for voice data.