Build SIP-WebSocket Audio Bridge
Budget: ₹75,000 – ₹150,000 INR
I want to move audio from a standard SIP trunk straight into LiveKit without using any third-party gateways. The bridge you build will sit between the SIP endpoint and a self-hosted LiveKit installation that I’ll be running on AWS. Once in place, every call that arrives over SIP should surface inside LiveKit as a clean WebSocket audio stream my existing WebRTC service can subscribe to in real time.
Scope and expectations
• Audio only: no video or text is required, but please handle common voice codecs (G.711 µ-law / A-law, Opus) and pass DTMF events cleanly.
• AWS deployment: the LiveKit stack will live in my AWS account, so the bridge must run happily there too—Docker or Kubernetes is fine as long as you supply the manifests or compose files.
• WebRTC integration: the finished bridge must expose an API or webhook so my current WebRTC app can create, join or end calls programmatically.
Deliverables
1. Source code for the SIP-to-WebSocket bridge with clear build instructions.
2. Infrastructure scripts (Terraform, Helm chart or similar) to spin up LiveKit and the bridge on AWS.
3. A short README covering configuration of the SIP trunk, expected environment variables and test steps.
4. End-to-end demo: place a SIP call, hear the audio inside a LiveKit room via WebSocket, and close the call without leaks.
Acceptance criteria
• Latency under 200 ms round-trip on a local AWS region test.
• No packet loss or distortion during a 30-minute call.
• Automatic reconnection if either the SIP trunk or WebSocket endpoint briefly drops.
If you’ve already worked with LiveKit, FreeSWITCH, Asterisk or similar tooling, that experience will help a lot. Let me know which approach you prefer for the SIP side—Go, C++, or Node are all acceptable—as long as the final bridge is reliable and well documented.
Scope and expectations
• Audio only: no video or text is required, but please handle common voice codecs (G.711 µ-law / A-law, Opus) and pass DTMF events cleanly.
• AWS deployment: the LiveKit stack will live in my AWS account, so the bridge must run happily there too—Docker or Kubernetes is fine as long as you supply the manifests or compose files.
• WebRTC integration: the finished bridge must expose an API or webhook so my current WebRTC app can create, join or end calls programmatically.
Deliverables
1. Source code for the SIP-to-WebSocket bridge with clear build instructions.
2. Infrastructure scripts (Terraform, Helm chart or similar) to spin up LiveKit and the bridge on AWS.
3. A short README covering configuration of the SIP trunk, expected environment variables and test steps.
4. End-to-end demo: place a SIP call, hear the audio inside a LiveKit room via WebSocket, and close the call without leaks.
Acceptance criteria
• Latency under 200 ms round-trip on a local AWS region test.
• No packet loss or distortion during a 30-minute call.
• Automatic reconnection if either the SIP trunk or WebSocket endpoint briefly drops.
If you’ve already worked with LiveKit, FreeSWITCH, Asterisk or similar tooling, that experience will help a lot. Let me know which approach you prefer for the SIP side—Go, C++, or Node are all acceptable—as long as the final bridge is reliable and well documented.