Skip to content

README.md

Latest commit

 

History

History
44 lines (32 loc) · 1.7 KB

README.md

File metadata and controls

44 lines (32 loc) · 1.7 KB

Study RAG: Flask + Gemini File Search App

A simple, clean study assistant web app built with Flask and Google Gemini.
You can upload your study notes (PDF, DOC, TXT, etc.) and then ask questions based on those notes.
The app sends your questions to Gemini and shows the answer in a nicely formatted way. Self-hosted the app using Pinggy.io. Check:https://pinggy.io/blog/how_to_use_file_search_tool_in_gemini_api_for_easy_rag_integration/

✨ Features

  • 📝 Upload notes (e.g., .txt, .pdf, .doc)
  • 🔎 Ask questions about your uploaded notes
  • 🧠 Uses Google Gemini (gemini-2.0-flash-lite-preview) for answering
  • 📚 File Search integration via Gemini File Search Store
  • ✅ Success message when a file is uploaded and indexed
  • ⚠ Friendly message when API quota (429 RESOURCE_EXHAUSTED) is hit
  • 🧾 Answer shown in clean paragraphs with basic formatting (Markdown → HTML)
  • 🎨 Simple & clean UI built with HTML + inline CSS

Application_demo1 Application_demo2

🧱 Tech Stack

  • Backend: Python, Flask
  • LLM / RAG: Google Gemini via google-genai SDK
  • Frontend: HTML (Jinja templates), minimal inline CSS
  • Template Engine: Jinja2 (Flask default)
  • Tunneling Tool: Pinggy (https://pinggy.io/)
  • Formatting: markdown Python library to render AI output nicely

📂 Project Structure

Example structure:

study-rag/
├─ app.py
├─ requirements.txt
├─ templates/
│  └─ index.html
└─ static/          # (optional, for images/icons if you add them later)