SIP-AI Voice Bridge Development
Budget: $250 – $750 USD
Build SIP-to-AI Voice Bridge (Inbound & Outbound Calls) – Python + Asterisk/FreeSWITCH
⸻
Budget:
Fixed price: $300 (non-negotiable)
⸻
Description:
We are looking for a developer to build a lightweight SIP interface that connects phone calls (inbound and outbound) to an AI assistant via WebSocket.
The AI engine and WebSocket endpoint are already functional.
We also have a dashboard UI and a working example of our API, which you will be expected to adapt and extend.
This is a focused integration task to bridge SIP audio with our AI infrastructure.
⸻
What We Need:
1. SIP Gateway Setup
• Use Asterisk or FreeSWITCH
• Handle both:
• Inbound SIP calls from trunk or softphone
• Outbound SIP calls triggered via REST API
• Provide basic configuration for testing (e.g. softphone, SIP trunk)
2. WebSocket Audio Bridge
• On call connect:
• Capture live RTP audio and stream it to our WebSocket endpoint (/ws/audio) using the openai-audio subprotocol
• Receive TTS audio responses and inject them back into the SIP call
• Handle call lifecycle: connect, hangup, timeout, and error states
3. REST API for Outbound Calls
• Create a POST endpoint /start-call to:
• Place an outbound SIP call to a specified number
• Automatically bridge the audio stream to the AI assistant
• We will provide a sample of our current API to extend
⸻
What We Already Have:
• A working AI backend (STT → GPT → TTS)
• A WebSocket audio endpoint
• A functional dashboard UI
• A working example API that needs adaptation
⸻
Technical Requirements:
• Python 3
• Runs in Docker
• You may use FastAPI, Flask, PJSIP, aiortc, or any reliable libraries
• Must support testing via softphone (Zoiper, Linphone, etc.)
• Code must be clean, well-documented, and modular
⸻
Deliverables:
• Dockerized SIP gateway and Python audio bridge
• Configuration for inbound and outbound SIP calls
• REST endpoint for call initiation
• Updated and working version of our example API
• README file with setup and testing instructions
⸻
Timeline:
• Start: Immediately
• Delivery: Within 7–10 days
⸻
Example Call Flows:
Inbound Call:
1. Incoming SIP call is received
2. Audio is streamed to AI via WebSocket
3. AI replies with audio, which is played back to the caller
Outbound Call:
1. API request is sent to /start-call
2. SIP call is placed
3. Audio is streamed both ways through the AI assistant
⸻
To Apply:
Please include:
• A short summary of your experience with SIP, Asterisk/FreeSWITCH, and Python
• Confirmation that you accept the $300 fixed price
• Estimated delivery timeline
We will provide:
• WebSocket documentation
• Sample API
• SIP credentials for testing
⸻
Budget:
Fixed price: $300 (non-negotiable)
⸻
Description:
We are looking for a developer to build a lightweight SIP interface that connects phone calls (inbound and outbound) to an AI assistant via WebSocket.
The AI engine and WebSocket endpoint are already functional.
We also have a dashboard UI and a working example of our API, which you will be expected to adapt and extend.
This is a focused integration task to bridge SIP audio with our AI infrastructure.
⸻
What We Need:
1. SIP Gateway Setup
• Use Asterisk or FreeSWITCH
• Handle both:
• Inbound SIP calls from trunk or softphone
• Outbound SIP calls triggered via REST API
• Provide basic configuration for testing (e.g. softphone, SIP trunk)
2. WebSocket Audio Bridge
• On call connect:
• Capture live RTP audio and stream it to our WebSocket endpoint (/ws/audio) using the openai-audio subprotocol
• Receive TTS audio responses and inject them back into the SIP call
• Handle call lifecycle: connect, hangup, timeout, and error states
3. REST API for Outbound Calls
• Create a POST endpoint /start-call to:
• Place an outbound SIP call to a specified number
• Automatically bridge the audio stream to the AI assistant
• We will provide a sample of our current API to extend
⸻
What We Already Have:
• A working AI backend (STT → GPT → TTS)
• A WebSocket audio endpoint
• A functional dashboard UI
• A working example API that needs adaptation
⸻
Technical Requirements:
• Python 3
• Runs in Docker
• You may use FastAPI, Flask, PJSIP, aiortc, or any reliable libraries
• Must support testing via softphone (Zoiper, Linphone, etc.)
• Code must be clean, well-documented, and modular
⸻
Deliverables:
• Dockerized SIP gateway and Python audio bridge
• Configuration for inbound and outbound SIP calls
• REST endpoint for call initiation
• Updated and working version of our example API
• README file with setup and testing instructions
⸻
Timeline:
• Start: Immediately
• Delivery: Within 7–10 days
⸻
Example Call Flows:
Inbound Call:
1. Incoming SIP call is received
2. Audio is streamed to AI via WebSocket
3. AI replies with audio, which is played back to the caller
Outbound Call:
1. API request is sent to /start-call
2. SIP call is placed
3. Audio is streamed both ways through the AI assistant
⸻
To Apply:
Please include:
• A short summary of your experience with SIP, Asterisk/FreeSWITCH, and Python
• Confirmation that you accept the $300 fixed price
• Estimated delivery timeline
We will provide:
• WebSocket documentation
• Sample API
• SIP credentials for testing