CustomerBot AI Backend

CustomerBot AI Backend is a multi-tenant backend for AI-powered customer support automation. It manages business workspaces, chat widgets, training documents, calls, subscriptions, integrations, and LangChain/LangGraph-based agent workflows.

The project is designed as the backend layer for a configurable customer-service platform that can ingest business knowledge and expose AI chat, voice, meeting scheduling, CRM, and support workflows.

Highlights

• Express.js API with a modular route/controller/service structure
• LangChain, LangGraph, OpenAI, and ChromaDB for AI agent workflows
• PostgreSQL persistence through Sequelize models and migrations
• Redis and Socket.IO support for real-time chat and state
• Training pipeline for documents, URLs, PDFs, and business data
• Multi-agent workflow modules for tool creation, supervisor routing, and meeting scheduling
• Integrations for HubSpot, Google Calendar/OAuth, Twilio, Stripe, email, Azure Blob Storage, S3, and Cloudinary
• Subscription, invitation, team group, call, chat widget, and business integration domains
• Docker support and database migrations/seeders

Tech Stack

• Node.js
• Express
• Sequelize
• PostgreSQL
• Redis
• Socket.IO
• LangChain / LangGraph
• OpenAI
• ChromaDB
• Stripe
• Twilio
• HubSpot API
• Google APIs
• AWS S3 / Azure Blob Storage

Architecture

src/
  controllers/          API controllers and AI workflow controllers
  routes/               REST endpoints for auth, chat, calls, training, billing, users
  integrations/         external services such as CRM, storage, calendar, Redis, email
  services/             domain services for subscriptions, pricing, training, summaries
  eventHandlers/        Socket.IO chat and user events
  validators/           request validation

migrations/             Sequelize database migrations
models/                 Sequelize models
config/                 database and environment configuration

Local Development

Install dependencies:

npm install

Start supporting services:

chroma run
redis-stack-server

Run migrations and seeders:

npx sequelize-cli db:migrate
npx sequelize-cli db:seed:all

Start the API:

npm run start:dev

Environment

Create a local .env file from .env.example and provide values for database, Redis, OpenAI, ChromaDB, JWT, Stripe, storage, email, CRM, and telephony integrations.

Security

Never commit real .env files, OAuth credential JSON, API keys, tokens, customer documents, or production exports. The repository is configured to ignore those files.