Patter: Custom Voice

Pipeline mode with independent STT and TTS providers -- use Deepgram for speech-to-text and ElevenLabs for text-to-speech, or mix and match any supported provider.

Part of the Patter Voice AI SDK.

Prerequisites

• Twilio account with a phone number
• OpenAI API key (used for the LLM)
• Deepgram API key for speech-to-text
• ElevenLabs API key for text-to-speech

Quick Start

Python

cd python
cp ../.env.example .env   # fill in your keys
pip install -r requirements.txt
python main.py

TypeScript

cd typescript
cp ../.env.example .env   # fill in your keys
npm install
npx tsx main.ts

Environment Variables

Variable	Required	Description
OPENAI_API_KEY	Yes	OpenAI API key (used for the LLM)
TWILIO_ACCOUNT_SID	Yes	Twilio account SID
TWILIO_AUTH_TOKEN	Yes	Twilio auth token
TWILIO_PHONE_NUMBER	Yes	Your Twilio phone number (E.164)
DEEPGRAM_API_KEY	Yes	Deepgram API key for STT
ELEVENLABS_API_KEY	Yes	ElevenLabs API key for TTS
WEBHOOK_URL	No	Public URL for webhooks (auto-tunneled if omitted)

What This Demonstrates

• provider="pipeline" for modular STT + LLM + TTS
• Patter.deepgram() for configuring Deepgram speech-to-text
• Patter.elevenlabs() for configuring ElevenLabs text-to-speech
• Custom STT/TTS selection independent of the LLM provider

Next Steps

• Inbound Agent — answer calls with AI
• Outbound Calls — place calls with AMD
• Tool Calling — add webhook-backed tools
• Full documentation
• All templates

License

MIT