Deploy and Host Chatty - Open-Source AI Customer Support Agent on Railway

Chatty is an open-source AI customer support platform trained on your store and documentation. It features autonomous 24/7 customer support, interactive Flow Builder, live agent takeover, WooCommerce store synchronization, meeting booking, and a voice agent interface.

About Hosting Chatty - Open-Source AI Customer Support Agent

Hosting Chatty on Railway deploys two coordinated services: a high-performance Python FastAPI backend (chatty-backend) and a Next.js 15 modern frontend web console (chatty-frontend). The architecture connects to your existing managed Supabase instance for Auth, PostgreSQL database, Realtime subscriptions, and Storage. Deploying on Railway gives you automated container healthchecks, public networking proxies, and seamless zero-downtime rollouts without managing underlying virtual machine infrastructure.

Why Deploy

Cost-effective and full data ownership: Full control over customer conversations and knowledge base data without expensive per-seat SaaS subscription fees.
Multi-channel integration: Connect website live chat, WhatsApp, Slack, and WooCommerce from a single unified dashboard.
Autonomous AI + Human takeover: AI answers routine customer questions instantly and smoothly escalates complex cases to human agents with full conversation context.
Fast deployment: Production-ready in minutes using managed Supabase and Google Gemini.

Common Use Cases

Autonomous e-commerce customer support with order status tracking and WooCommerce product catalog sync.
24/7 AI-powered website live chat widget with automated lead capture and qualification.
Visual flow-based workflow automation and interactive voice support agents.
Hybrid human-in-the-loop support with instant agent takeover and customer context inspection.

Dependencies for

Supabase Project: Provides PostgreSQL database, Authentication, Realtime, and Storage.
Database Migrations: Applied to your Supabase PostgreSQL instance using python apply_migrations.py.
Google Gemini API Key: Used for AI response generation and document embeddings.