AI Search

This is a fairly simple project that was inspired by Perplexity. It uses llama-3.3-70b-versatile (hosted on Groq) and Google Custom Search JSON API.

Tech Stack

Backend

• Node.js
• Express.js
• No TypeScript (yet)

Frontend

• React
• TypeScript
• Tailwind CSS (& CSS)

Development

1. Clone the repo.
2. Make sure Node.js is installed. I've tested with v22.
3. Run npm install in both backend and frontend subdirs.
4. Create a .env file in backend. Check backend/src/env.js for variables. Also make sure to set PORT to 8080.
5. Run npm run dev in both subdirs.

Acknowledgements

1. Perplexity for inspiration
2. Meta for Llama
3. Groq for inference
4. Google for data
5. bolt and v0 for frontend
6. Claude for helping me with prompts
7. And ChatGPT of course