selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

GitHub
AI Native Call Center logo

Self-host AI Native Call Center

v0.7.1

AI-native call center where AI answers every call first

BetaOpen SourceApache-2.0
Updated4 hours ago
LanguageGo
Setup difficultyAdvanced
Stars
441
CPU
4 CPUs
RAM
4 GiB
Ports
8080 (HTTP API, event stream and web interface), 5060 (SIP signalling (UDP/TCP)), 5066 (Browser phone WebSocket transport), 6060 (AI bot SIP endpoint), 18021 (FreeSWITCH ESL connection), 9090 (Metrics listener (/metrics, /healthz, /readyz)), 8090 (Transcription audio stream listener), 40000 (AI bot RTP media range start), 40999 (AI bot RTP media range end), 16384 (FreeSWITCH RTP media range start)
Requires
FreeSWITCH, PostgreSQL
Docker
rasonyang/ai-native-callcenter
Install with Docker Compose
Agent desktop showing inbound call answered, live transcript filling in, then after-call work

What is AI Native Call Center?

AI Native Call Center is a self-hosted call center platform where AI voice agents answer every incoming call by default. The AI leg runs as a SIP endpoint inside the Go application, bridged by FreeSWITCH, with audio passed directly to realtime speech-to-speech providers (OpenAI, Grok, Qwen, Doubao, or a compatible gateway). A flow engine steers conversations through phases and tools, and transfers to human agents in mod_callcenter queues when needed — the agent's browser screen pops with the transcript before the phone rings. Agents use a companion Chrome extension (web-sip-phone) for audio. Every call produces a single CDR with transcript and recording, stored on the filesystem or any S3-compatible backend. The stack runs as three containers (PostgreSQL 18, FreeSWITCH 1.11, and the Go application with embedded SPA) via Docker Compose, with a one-line installer for Linux and macOS. Minimum 4 CPUs and 4 GiB RAM. Licensed Apache-2.0.

Should you self-host AI Native Call Center?

AI Native Call Center scores 48/100 on selfhost.directory (how scoring works) and ranks #10 of 14 in Customer Support. It has 441 GitHub stars.

It shipped 9 stable releases in the past year. The latest, v0.7.1, came out on 28 October 2026. New versions land roughly every 5 days.

Setup is advanced: plan for several services and some ongoing maintenance. An official Docker image is available. It is Apache-2.0-licensed and written mostly in Go.

Its closest alternative on selfhost.directory is chatwoot (84/100, 38 470 stars). It scores higher overall, so compare the two before you pick. See AI Native Call Center vs chatwoot side by side.

Features

  • AI Voice Agents
  • Bilingual support
  • Call Queues
  • Call Recording
  • CDR & Webhooks
  • Chrome Extension PhoneSeparate repository, installed from Chrome Web Store
  • Flow Engine
  • Human handoff
  • Live transcriptionOptional, requires additional configuration
  • Multiple Voice ProvidersOne provider per deployment, chosen at startup
  • Outbound AI Campaigns
  • Post-call TranscriptionOptional, requires Qwen ASR
  • REST API
  • Server-Sent Events
  • Web Agent Desktop

AI Native Call Center README

How to install AI Native Call Center

Install AI Native Call Center with Docker Compose

Environment variables
AICC_PROVIDER, OPENAI_API_KEY/XAI_API_KEY/ALIYUN_API_KEY/DOUBAO_API_KEY/REALTIME_API_KEY, FS_EXTERNAL_IP, POSTGRES_PASSWORD, ESL_PASSWORD, LUA_PASSWORD, AICC_SEED_PASSWORD

One-line installer for Linux, installs to /opt/aicc with host networking

AI Native Call Center releases & version history

More in VoIP

Rate this project

Click to rate — no account needed.

0 comments

Posting as guest
Loading comments…