it0/deploy/docker/docker-compose.yml

name: it0

services:
  # ===== Infrastructure =====
  postgres:
    image: postgres:16-alpine
    container_name: it0-postgres
    restart: unless-stopped
    environment:
      POSTGRES_USER: ${POSTGRES_USER:-it0}
      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-it0_dev}
      POSTGRES_DB: ${POSTGRES_DB:-it0}
    ports:
      - "15432:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data
      - ./scripts/init-databases.sh:/docker-entrypoint-initdb.d/init-databases.sh
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER:-it0}"]
      interval: 10s
      timeout: 5s
      retries: 5
    networks:
      - it0-network

  redis:
    image: redis:7-alpine
    container_name: it0-redis
    restart: unless-stopped
    ports:
      - "16379:6379"
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 10s
      timeout: 5s
      retries: 5
    networks:
      - it0-network

  # ===== API Gateway =====
  api-gateway:
    build:
      context: ../../packages/gateway
    container_name: it0-api-gateway
    restart: unless-stopped
    environment:
      - JWT_SECRET=${JWT_SECRET:-dev-jwt-secret}
    ports:
      - "18000:8000"
      - "18001:8001"
    depends_on:
      auth-service:
        condition: service_healthy
      agent-service:
        condition: service_healthy
      ops-service:
        condition: service_healthy
      inventory-service:
        condition: service_healthy
      monitor-service:
        condition: service_healthy
      comm-service:
        condition: service_healthy
      audit-service:
        condition: service_healthy
      billing-service:
        condition: service_healthy
      version-service:
        condition: service_healthy
      presence-service:
        condition: service_healthy
      referral-service:
        condition: service_healthy
      notification-service:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "kong", "health"]
      interval: 10s
      timeout: 5s
      retries: 3
    networks:
      - it0-network

  # ===== Backend Services (shared Dockerfile.service) =====
  auth-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: auth-service
        SERVICE_PORT: 3001
    container_name: it0-auth-service
    restart: unless-stopped
    ports:
      - "13001:3001"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - JWT_SECRET=${JWT_SECRET:-dev-jwt-secret}
      - JWT_REFRESH_SECRET=${JWT_REFRESH_SECRET:-dev-jwt-refresh-secret}
      - AUTH_SERVICE_PORT=3001
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3001/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks:
      - it0-network

  agent-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: agent-service
        SERVICE_PORT: 3002
    container_name: it0-agent-service
    restart: unless-stopped
    ports:
      - "13002:3002"
    volumes:
      - ${HOME}/.claude:/home/appuser/.claude
      - ${HOME}/.claude.json:/home/appuser/.claude.json
      - claude_tenants:/data/claude-tenants
      - ${HOME}/.ssh/rwadurian_ed25519:/tmp/host-ssh-key:ro
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - AGENT_ENGINE_TYPE=claude_agent_sdk
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
      - ANTHROPIC_BASE_URL=${ANTHROPIC_BASE_URL}
      - OPENAI_API_KEY=${OPENAI_API_KEY}
      - OPENAI_BASE_URL=${OPENAI_BASE_URL}
      - AGENT_SERVICE_PORT=3002
      - INVENTORY_SERVICE_URL=http://inventory-service:3004
      - INTERNAL_API_KEY=${INTERNAL_API_KEY:-changeme-internal-key}
      - VAULT_MASTER_KEY=${VAULT_MASTER_KEY:-dev-vault-key}
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3002/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks:
      - it0-network

  ops-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: ops-service
        SERVICE_PORT: 3003
    container_name: it0-ops-service
    restart: unless-stopped
    ports:
      - "13003:3003"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - OPS_SERVICE_PORT=3003
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3003/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks:
      - it0-network

  inventory-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: inventory-service
        SERVICE_PORT: 3004
    container_name: it0-inventory-service
    restart: unless-stopped
    ports:
      - "13004:3004"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - VAULT_MASTER_KEY=${VAULT_MASTER_KEY:-dev-vault-key}
      - INTERNAL_API_KEY=${INTERNAL_API_KEY:-changeme-internal-key}
      - INVENTORY_SERVICE_PORT=3004
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3004/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
    networks:
      - it0-network

  monitor-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: monitor-service
        SERVICE_PORT: 3005
    container_name: it0-monitor-service
    restart: unless-stopped
    ports:
      - "13005:3005"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - MONITOR_SERVICE_PORT=3005
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3005/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
    networks:
      - it0-network

  comm-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: comm-service
        SERVICE_PORT: 3006
    container_name: it0-comm-service
    restart: unless-stopped
    ports:
      - "13006:3006"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - TWILIO_ACCOUNT_SID=${TWILIO_ACCOUNT_SID}
      - TWILIO_AUTH_TOKEN=${TWILIO_AUTH_TOKEN}
      - TWILIO_PHONE_NUMBER=${TWILIO_PHONE_NUMBER}
      - COMM_SERVICE_PORT=3006
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3006/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks:
      - it0-network

  audit-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: audit-service
        SERVICE_PORT: 3007
    container_name: it0-audit-service
    restart: unless-stopped
    ports:
      - "13007:3007"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - AUDIT_SERVICE_PORT=3007
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3007/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
    networks:
      - it0-network

  billing-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: billing-service
        SERVICE_PORT: 3010
    container_name: it0-billing-service
    restart: unless-stopped
    ports:
      - "13010:3010"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - BILLING_SERVICE_PORT=3010
      - STRIPE_SECRET_KEY=${STRIPE_SECRET_KEY}
      - STRIPE_WEBHOOK_SECRET=${STRIPE_WEBHOOK_SECRET}
      - ALIPAY_APP_ID=${ALIPAY_APP_ID}
      - ALIPAY_PRIVATE_KEY=${ALIPAY_PRIVATE_KEY}
      - WECHAT_MCH_ID=${WECHAT_MCH_ID}
      - WECHAT_API_KEY_V3=${WECHAT_API_KEY_V3}
      - WECHAT_APP_ID=${WECHAT_APP_ID}
      - COINBASE_COMMERCE_API_KEY=${COINBASE_COMMERCE_API_KEY}
      - COINBASE_COMMERCE_WEBHOOK_SECRET=${COINBASE_COMMERCE_WEBHOOK_SECRET}
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3010/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks:
      - it0-network

  version-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: version-service
        SERVICE_PORT: 3009
    container_name: it0-version-service
    restart: unless-stopped
    ports:
      - "13009:3009"
    volumes:
      - version_data:/data/versions
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - VERSION_SERVICE_PORT=3009
      - DOWNLOAD_BASE_URL=https://it0api.szaiai.com/downloads/versions
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3009/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
    networks:
      - it0-network

  presence-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: presence-service
        SERVICE_PORT: 3011
    container_name: it0-presence-service
    restart: unless-stopped
    ports:
      - "13011:3011"
    environment:
      - DATABASE_URL=postgresql://${POSTGRES_USER:-it0}:${POSTGRES_PASSWORD:-it0_dev}@postgres:5432/${POSTGRES_DB:-it0}
      - REDIS_HOST=redis
      - REDIS_PORT=6379
      - REDIS_DB=10
      - APP_PORT=3011
      - JWT_SECRET=${JWT_SECRET:-dev-jwt-secret}
      - PRESENCE_WINDOW_SECONDS=300
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3011/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks:
      - it0-network

  referral-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: referral-service
        SERVICE_PORT: 3012
    container_name: it0-referral-service
    restart: unless-stopped
    ports:
      - "13012:3012"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - REDIS_URL=redis://redis:6379
      - REFERRAL_SERVICE_PORT=3012
      - JWT_SECRET=${JWT_SECRET:-dev-jwt-secret}
      - INTERNAL_API_KEY=${INTERNAL_API_KEY:-changeme-internal-key}
      - APP_REFERRAL_BASE_URL=https://it0api.szaiai.com
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3012/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks:
      - it0-network

  notification-service:
    build:
      context: ../..
      dockerfile: Dockerfile.service
      args:
        SERVICE_NAME: notification-service
        SERVICE_PORT: 3013
    container_name: it0-notification-service
    restart: unless-stopped
    ports:
      - "13013:3013"
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_USERNAME=${POSTGRES_USER:-it0}
      - DB_PASSWORD=${POSTGRES_PASSWORD:-it0_dev}
      - DB_DATABASE=${POSTGRES_DB:-it0}
      - NOTIFICATION_SERVICE_PORT=3013
      - JWT_SECRET=${JWT_SECRET:-dev-jwt-secret}
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://localhost:3013/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      postgres:
        condition: service_healthy
    networks:
      - it0-network

  # ===== LiveKit Infrastructure =====
  # NOTE: livekit-server, voice-agent, voice-service use host networking
  # to eliminate docker-proxy overhead for real-time audio (WebRTC UDP).
  # Bridge mode created 600+ docker-proxy processes for port-range mappings.
  livekit-server:
    image: livekit/livekit-server:latest
    container_name: it0-livekit-server
    restart: unless-stopped
    network_mode: host
    command: --config /etc/livekit.yaml
    volumes:
      - ./livekit.yaml:/etc/livekit.yaml:ro
    healthcheck:
      test: ["CMD-SHELL", "wget --no-verbose --tries=1 --spider http://localhost:7880 || exit 1"]
      interval: 10s
      timeout: 5s
      retries: 3

  voice-agent:
    build:
      context: ../../packages/services/voice-agent
    container_name: it0-voice-agent
    restart: unless-stopped
    network_mode: host
    volumes:
      - ../../data/voice-models/huggingface:/root/.cache/huggingface
      - ../../data/voice-models/torch-hub:/root/.cache/torch/hub
    environment:
      - LIVEKIT_URL=ws://127.0.0.1:7880
      - LIVEKIT_API_KEY=devkey
      - LIVEKIT_API_SECRET=devsecret
      - AGENT_SERVICE_URL=http://127.0.0.1:13002
      - STT_PROVIDER=${STT_PROVIDER:-local}
      - TTS_PROVIDER=${TTS_PROVIDER:-local}
      - WHISPER_MODEL=${WHISPER_MODEL:-base}
      - WHISPER_LANGUAGE=${WHISPER_LANGUAGE:-zh}
      - KOKORO_VOICE=${KOKORO_VOICE:-zf_xiaoxiao}
      - DEVICE=${VOICE_DEVICE:-cpu}
      - OPENAI_API_KEY=${OPENAI_API_KEY}
      - OPENAI_BASE_URL=${OPENAI_BASE_URL}
      - OPENAI_STT_MODEL=${OPENAI_STT_MODEL:-gpt-4o-transcribe}
      - OPENAI_TTS_MODEL=${OPENAI_TTS_MODEL:-gpt-4o-mini-tts}
      - OPENAI_TTS_VOICE=${OPENAI_TTS_VOICE:-coral}
    depends_on:
      livekit-server:
        condition: service_healthy
      agent-service:
        condition: service_healthy

  # ===== Voice Service (LiveKit token + Twilio) =====
  # NOTE: voice-service stays on bridge networking (single TCP port, no proxy overhead).
  # Only livekit-server and voice-agent need host mode (UDP port ranges).
  voice-service:
    build:
      context: ../../packages/services/voice-service
    container_name: it0-voice-service
    restart: unless-stopped
    ports:
      - "13008:3008"
    volumes:
      - ../../data/voice-models/huggingface:/root/.cache/huggingface
      - ../../data/voice-models/torch-hub:/root/.cache/torch/hub
    environment:
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
      - ANTHROPIC_BASE_URL=${ANTHROPIC_BASE_URL}
      - AGENT_SERVICE_URL=http://agent-service:3002
      - WHISPER_MODEL=${WHISPER_MODEL:-base}
      - KOKORO_MODEL=${KOKORO_MODEL:-kokoro-82m}
      - KOKORO_VOICE=${KOKORO_VOICE:-zf_xiaoxiao}
      - DEVICE=${VOICE_DEVICE:-cpu}
      - STT_PROVIDER=${STT_PROVIDER:-local}
      - TTS_PROVIDER=${TTS_PROVIDER:-local}
      - OPENAI_API_KEY=${OPENAI_API_KEY}
      - OPENAI_BASE_URL=${OPENAI_BASE_URL}
      - OPENAI_STT_MODEL=${OPENAI_STT_MODEL:-gpt-4o-transcribe}
      - OPENAI_TTS_MODEL=${OPENAI_TTS_MODEL:-gpt-4o-mini-tts}
      - OPENAI_TTS_VOICE=${OPENAI_TTS_VOICE:-coral}
      - LIVEKIT_API_KEY=devkey
      - LIVEKIT_API_SECRET=devsecret
      - LIVEKIT_WS_URL=ws://14.215.128.96:7880
    healthcheck:
      test: ["CMD-SHELL", "python3 -c \"import urllib.request; urllib.request.urlopen('http://localhost:3008/docs')\""]
      interval: 30s
      timeout: 10s
      retries: 5
      start_period: 120s
    depends_on:
      - agent-service
    networks:
      - it0-network

  # ===== Frontend =====
  web-admin:
    build:
      context: ../../it0-web-admin
    container_name: it0-web-admin
    restart: unless-stopped
    ports:
      - "13000:3000"
    environment:
      - HOSTNAME=0.0.0.0
      - API_BASE_URL=http://api-gateway:8000
      - NEXT_PUBLIC_API_BASE_URL=/api/proxy
      - NEXT_PUBLIC_WS_URL=wss://it0api.szaiai.com
    healthcheck:
      test: ["CMD-SHELL", "node -e \"require('http').get('http://127.0.0.1:3000/',r=>{process.exit(r.statusCode<500?0:1)}).on('error',()=>process.exit(1))\""]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 15s
    depends_on:
      - api-gateway
    networks:
      - it0-network

volumes:
  postgres_data:
  claude_tenants:
  version_data:

networks:
  it0-network:
    driver: bridge