Starter Folder Structure spur-ai-chat/ ├── backend/ │ ├── src/ │ │ ├── routes/ │ │ │ └── chat.ts │ │ ├── services/ │ │ │ ├── chat.service.ts │ │ │ └── llm.service.ts │ │ ├── db/ │ │ │ ├── prisma.ts │ │ │ └── schema.prisma │ │ ├── utils/ │ │ │ └── validate.ts │ │ └── server.ts │ ├── .env.example │ ├── package.json │ └── tsconfig.json │ ├── frontend/ │ ├── src/ │ │ ├── lib/api.ts │ │ ├── components/Chat.svelte │ │ └── routes/+page.svelte │ ├── package.json │ └── svelte.config.js │ └── README.md

spur-ai-live-chat

A mini AI-powered live chat support agent built with TypeScript, Svelte, and a real LLM, simulating a production-ready customer support workflow.

Spur AI Live Chat Agent

This project is a mini AI-powered customer support chat widget built as a take-home assignment for the Founding Full-Stack Engineer role at Spur.

Tech Stack

• Backend: Node.js, TypeScript, Express
• Frontend: SvelteKit
• Database: PostgreSQL (Prisma)
• LLM: OpenAI
• Deployment: Render + Vercel

Features

• Live chat UI with typing indicator
• Session-based conversation persistence
• Chat history restored on reload
• Real LLM-powered responses
• Robust input validation and error handling

Running Locally

Backend

cd backend
npm install
cp .env.example .env
npx prisma migrate dev
npm run dev

### Frontend
```bash
cd frontend
npm install
npm run dev
OPENAI_API_KEY=your_key
DATABASE_URL=postgresql://...
LLM Notes

Provider: OpenAI

Prompt includes store FAQ + guardrails

Last messages included for context

Token limits applied for cost control

Trade-offs & Future Improvements

Streaming responses

Redis for session caching

Tool calling for order status / refunds

Admin dashboard

Multi-channel support