GenAI

A simple Retrieval-Augmented Generation (RAG) API for asking questions about uploaded documents. Uses FastAPI, Google Gemini and Chroma vector database.

✨ Features

• Upload multiple files (PDF, DOCX, TXT, JPG , PNG, CSV, DB)
• Contexts are generated with overlap
• Ask questions
• Relevant context and question will be sent to the LLM(Gemini)
• Get answers with exact source and context

❌ Cons

• Once you reload the page , the messages are gone. No way to retrieve them back.
• The Gemini free limit may cause a temporary intervention.

Installation

pip install -r requirements.txt

Quick Start/Run

For development(locally): fastapi dev main.py

How to use it?

• Upload all the files at once at the very beginning.
• Ask questions.

Enviroment Variables(Required)

GEMINI_API_KEY=your_gemini_api_key

This must be present in your environment variables.

API Documentation

1. ('/') -> returns index.html
2. ('/query') -> returns query.html
3. ('/api/query') -> makes a POST request to the backend with user's questions and 
                    gets the response in return

Sample Folders & Files

1. 'files' directory contains the sample files.
2. 'uploads' directory stores the uploaded documents.

Screenshots

Index page

Query page