F1 RAG Chatbot

F1 RAG Chatbot is a minimal Formula 1 assistant built with Next.js. It lets fans ask questions and get concise answers, styled with a Ferrari‑inspired UI.

Tech Stack

• Next.js App Router (TypeScript)
• OpenAI API (gpt‑4o for answers, text‑embedding‑3‑small for retrieval)
• DataStax Astra DB for vector search over scraped F1 content

Quick Start

npm install
npm run dev
# open http://localhost:3000

Add env vars in .env.local:

OPENAI_API_KEY=your_key_here
ASTRA_DB_API_ENDPOINT=...
ASTRA_DB_APPLICATION_TOKEN=...
ASTRA_DB_NAMESPACE=...
ASTRA_DB_COLLECTION=f1gpt

To ingest data into Astra DB:

npm run seed

Project Structure

• app/page.tsx – chat UI (messages, input, suggestions)
• app/api/chat/route.ts – server endpoint calling OpenAI
• app/components/* – UI building blocks (bubbles, loader, suggestions)
• app/global.css – theme and layout
• scripts/loadDb.ts – scraping/embedding loader for Astra DB

Notes

• Uses non‑streaming responses for simplicity
• RAG context is stored in Astra DB; the assistant runs on gpt‑4o

License

MIT