This is an AI-powered chatbot prototype built as a modular boilerplate for business-ready deployments. It uses a fine-tuned OpenAI model and supports a standalone frontend/backend architecture.
- Frontend: Built with React + TailwindCSS
- Backend: Node.js + Express
- AI: Fine-tuned GPT model via OpenAI API
- Environment:
.env-based configuration - Modular: Designed to be embedded into other websites
chatbot-demo/
├── .env # Global environment config (API keys, model ID)
├── backend/ # Express API server
│ ├── index.js
│ └── ...
├── frontend/ # React frontend
│ ├── src/
│ └── ...
├── data/ # Fine-tuning training files
├── .gitignore
├── package.json
└── README.md# From project root
npm install
cd backend && npm install
cd ../frontend && npm install# .env at root
OPENAI_API_KEY=your-api-key
FINE_TUNED_MODEL_ID=ft:gpt-4o:your-org::your-model
OPENAI_ORG_ID=your-org-id (optional)In separate terminals:
# Terminal 1 (backend)
cd backend
npm start
# Terminal 2 (frontend)
cd frontend
npm run devApp should be running at http://localhost:3000 (or whatever port Vite is configured for).
- Chat interface with OpenAI fine-tuned model
- Backend toggles between default and custom model
- Embeddable, standalone frontend component (WIP)
- Developer-friendly structure for reuse & extension
# Frontend
cd frontend
npm run dev # start Vite dev server
npm run build # build for production
# Backend
cd backend
npm start # start Express server| Key | Description |
|---|---|
OPENAI_API_KEY |
Required. Your OpenAI API Key |
FINE_TUNED_MODEL_ID |
Optional. Your fine-tuned model ID |
OPENAI_ORG_ID |
Optional. Your OpenAI org ID |
- Embed frontend as a drop-in widget
- Multi-session support
- Admin interface for model config
- Deployment pipeline setup
Chris Brown — cbreazy.com
GitHub: @CBreazy
This is a private repository. Not licensed for public or commercial use. All rights reserved © Brown Creative Company