RAG Chatbot

This is a small test project built to get familiar with how RAG, embeddings, and LangChain work in practice.

The app lets you:

• embed the contents of a website URL into an in-memory vector store
• inspect what gets saved after embedding
• ask questions about the embedded websites using retrieval augmented generation

It is intentionally simple and is meant as a learning playground more than a production-ready application.

Getting Started

Run the development server:

pnpm dev

Then open http://localhost:3000 in your browser.

You will need an OpenAI API key, either:

• set as OPENAI_API_KEY
• or entered in the UI when embedding and chatting

Tech Stack

• Next.js
• LangChain
• OpenAI embeddings and chat models
• in-memory MemoryVectorStore for retrieval

Purpose

The goal of this repo is to experiment with the core pieces of a RAG workflow:

1. load content from a URL
2. split it into chunks
3. create embeddings
4. store those embeddings in a vector database
5. retrieve relevant chunks for a user question
6. generate an answer grounded in the retrieved context