🧠 DocMind — Chat with Your Documents

Upload any PDF and ask questions in natural language. DocMind retrieves the exact answer from your document with source citations.

✨ Features

• Multi-turn conversation with memory
• Source citations with page numbers
• Multiple PDF support
• Premium dark UI
• Completely free — no paid APIs

🏗️ Architecture

PDF → Text Chunks (500 tokens, 50 overlap) → Gemini Embeddings → FAISS Vector Store → Retriever (top-4 chunks) → Gemini 2.5 Flash → Answer + Sources

🛠️ Tech Stack

Layer	Technology
LLM	Google Gemini 2.5 Flash
Embeddings	Google Generative AI Embeddings
Vector Store	FAISS
Framework	LangChain
UI	Streamlit
PDF Parsing	PyPDF

🚀 Run Locally

1. Clone the repo

   git clone https://github.com/k-charansai/docmind.git
   cd docmind

2. Install dependencies

   pip install -r requirements.txt

3. Add your API key Create a .env file:

   GOOGLE_API_KEY=your_key_here

   Get a free key at: aistudio.google.com

4. Run the app

   streamlit run app.py

💡 How to Use

1. Upload one or more PDF files using the sidebar
2. Click "Process Documents" and wait for embedding
3. Type your question in the chat input
4. Ask follow-up questions — DocMind remembers context
5. Click "Sources" under any answer to see exact snippets

⚠️ Rate Limits

Free tier allows 100 embedding requests per minute. Large PDFs (200+ chunks) will process in batches automatically.

🔮 Future Improvements

• Support for DOCX and TXT files
• Streaming LLM responses
• User authentication
• Cloud vector store (Pinecone)
• Export conversation as PDF