This guide outlines how to enable and use Exotel’s Voice Streaming capabilities via the Voicebot Applet or Stream Applet. These applets allow real-time audio streaming between your customer calls and bot platforms over WebSocket.
Step 1: Account Setup
Before enabling streaming services, ensure you have an active Exotel account in the appropriate region.
Choose the Right Instance
Note: IP–PSTN intermixing is supported only on the Mumbai (Veeno) instance.
Step 2: Complete KYC
Voice streaming access requires completion of your KYC verification.
Submit your company documentation as per standard onboarding.
- Refer to the complete KYC process here: How to complete KYC for Exotel
- Wait for confirmation from the Exotel compliance team before proceeding.
Note: KYC and streaming enablement can happen in parallel.
Step 3: Enable Streaming Features
Raise a request to enable the required streaming applets.
Raise a Support Ticket
Send an email to hello@exotel.com with the subject line: Enable Stream/Voicebot Applet for [Your Account SID] Include a brief note on your use case. CC your Account Director (AD) or Customer Success Manager (CSM).(No worries, if you don’t know your AD and CSM, our customer happiness team will add them.)
From a compliance point of view, we monitor and approve this in conjunction with your KYC.
Step 4: Integration and Support
Once access is enabled, proceed with integration using the documentation below.
Documentation Links
- Voicebot & Stream Applet Integration: https://support.exotel.com/support/solutions/articles/3000108630-working-with-the-stream-and-voicebot-applet
- Streaming Passthru Parameters Handling: Passthru Applet for Voice Streaming (Voicebot and Stream Applets)
Support Process
- For any integration-related support, your Customer Success Manager (CSM) will assist.
- Technical issues or escalations should be directed to hello@exotel.com.
Capabilities Overview
- Real-time audio streaming over WebSocket (WSS)
- Voicebot Applet: Bidirectional audio interaction
- Stream Applet: Unidirectional audio streaming
- Passthru Applet: Stream lifecycle metadata, error reporting, and routing context
- SIP Trunk fallback and external call routing available via Passthru (Mumbai region only)
- Supports major bot platforms.
Best Practices
Always choose the Mumbai instance for IP–PSTN or SIP hybrid integrations.
Disconnect the WebSocket from your bot when the conversation is complete to allow the call flow to proceed to the next applet.
Ensure media chunking follows Exotel's recommended size (typically 100ms PCM chunks with 3200 bytes of raw audio).
Use the Passthru Applet after Voicebot to capture stream completion, errors, and call routing metadata.
Avoid long delays in establishing WebSocket connections to prevent streaming failure or fallback.
Use dynamic WebSocket URLs when handling multi-tenant or user-segmented bot sessions.
Monitor StreamSID, Duration, Status, and DisconnectedBy via Passthru for complete lifecycle debugging.
Glossary
- WSS (WebSocket Secure): Secure protocol used for encrypted audio streaming.
- Voicebot Applet: Enables bidirectional audio exchange between the bot and the caller.
- Stream Applet: Allows unidirectional audio capture from the caller to your bot.
- Passthru Applet: Captures stream metadata, errors, and acts as a decision trigger for routing.
- StreamSID: Unique identifier for each stream session, useful for logging and tracking.
Common Use Cases
- Voicebot handling FAQ or collections for fintech and NBFCs
- Call deflection bots for first-level support in marketplaces
- Healthcare appointment management via AI bots
- Bidirectional streaming for GenAI-based interview bots
- Smart routing from bot to agent based on sentiment or keywords
Final Notes
Once these steps are completed and access is approved, you will be ready to stream calls to your own bot infrastructure in real time and build powerful conversational workflows using Exotel’s infrastructure.
For questions, assistance, or deployment readiness, contact your CSM or raise a support request via hello@exotel.com.