/sql-chatbot

chat with sql databases

Primary LanguageTypeScript

RAG with web pages. By Yago López

  • This proof of concept allows to make questions about the contents of a web page and get answers

  • It uses RAG: Retrieval Augmented Generation

    • In-memory Vector Database
    • Mistral Embeddings: to create vectorial representations of unstructured information (text)
    • Similarity Search
    • Mistral LLM

  • The user can provide the URL to the web page

  • The answer is scoped to the web page text. . If the question is not related to the web page the AI will not be able to provide a response. This avoid hallucinations that happen with general chat bots

Demo

Getting Started

First, run the development server:

npm run dev
# or
yarn dev
# or
pnpm dev
# or
bun dev
  • Open http://localhost:3000 with your browser to see the result.
  • Select or Add a URL in the input text box
  • Type a question about the text of the web page in the input text box
  • Submit the question

Technical Stack

  • LLM: Mistral AI
  • AI Frameworks:
    • Vercel AI SDK
    • LangChain
  • Frontend Framework: NextJS
  • UI Library: Shadcdn
  • Typescript

Environment Variables Needed

MISTRAL_API_KEY=
LANGCHAIN_TRACING_V2="true"
LANGCHAIN_API_KEY=

(Not provided in the project)