Hi AG,
Thanks — we’re interested in learning more about your self‑hosted build. A few quick notes and next steps:
- Currently Vapi does not offer a general self‑hosting/Docker image for all customers; on‑prem deployments are available for large enterprises and via AWS Marketplace in limited cases (see related threads).
- If you want to prototype a custom stack that connects to Vapi (e.g., custom LLM or local telemetry), the docs show how to run local servers and expose them with ngrok:
https://docs.vapi.ai/server-url/developing-locally and how to wire a custom LLM endpoint:
https://docs.vapi.ai/customization/custom-llm/using-your-server.
- If you’ve built a one‑click Docker deployment that fully replaces Vapi (or mirrors its orchestration), please share:
1. repo or Docker image URL and deployment instructions
2. which Vapi features are implemented (STT/TTS providers, signaling, orchestration, webhook handling, billing telemetry, security/compliance)
3. any licensing or third‑party provider requirements (OpenAI keys, Deepgram, LiveKit, Pipecat etc.)
Once you share the repo/image and a short feature matrix I can review compatibility gaps and advise on integration points or what’s needed to match Vapi functionality.