Smart Academic Assistant

AI-powered academic assistant using RAG (Retrieval-Augmented Generation). Upload study materials in PDF format and get instant answers to your questions using Google Gemini, LangChain, HuggingFace embeddings, and ChromaDB.

Features

• Upload PDF documents (notes, textbooks, study materials)
• Automatically create vector-based knowledge base
• Ask natural language questions about content
• Context-aware answers using Gemini model
• Simple Streamlit web interface

Tech Stack

• Python - Core development
• Streamlit - Web interface
• LangChain - RAG framework
• ChromaDB - Vector database
• HuggingFace Sentence Transformers - Embeddings
• Google Generative AI Gemini - Language model

Installation

Setup

# Install dependencies
pip install -r requirements.txt

# Set Google credentials
export GOOGLE_APPLICATION_CREDENTIALS="/path/to/your/credentials.json"

Run Application

streamlit run app.py

Authentication

Uses Google JSON credentials file instead of API key. Set the environment variable:

GOOGLE_APPLICATION_CREDENTIALS="/absolute/path/to/your/credentials.json"

Use Cases

• Extract specific information from student lists (birthdays, phone numbers, exam dates)
• Query uploaded lecture notes with specific questions
• Review PDF study guides through summary questions
• Search through academic documents and textbooks

Limitations

• Works best with well-formatted text PDFs
• Raw unstructured data may require proper formatting
• Document chunking might split relevant information
• Performance depends on PDF text quality

License

MIT License