🤖 PDF Chatbot AI — A Generative AI Application

A Generative AI-powered chatbot built with Streamlit that allows users to upload PDF documents and get detailed, human-like answers to their questions using state-of-the-art language models.

🚀 Features

✅ Upload and analyze PDF files 📄
✅ Extract and segment text using PyPDF2 + NLTK ✂️
✅ Convert text into semantic embeddings using all-mpnet-base-v2 🔍
✅ Retrieve the most relevant context based on user questions 🧠
✅ Generate context-aware answers using FLAN-T5-Large (Generative AI) ✨
✅ Clean and interactive UI with Streamlit 💬

🛠️ Installation

1⃣ Clone the repository

git clone https://github.com/shantanumohod23/pdf-chatbot-ai.git
cd pdf-chatbot-ai

2⃣ Create and activate a virtual environment

# Windows
python -m venv .venv
.venv\Scripts\activate

# Mac/Linux
python3 -m venv .venv
source .venv/bin/activate

3⃣ Install dependencies

pip install -r requirements.txt

▶️ How to Run

Once installed, start the Streamlit app:

streamlit run app.py

Upload a PDF file, ask your question, and let the AI generate an insightful answer! 🤖📚

📁 Folder Structure

📂 pdf-chatbot-ai
 ├── app.py            # Main Streamlit app
 ├── requirements.txt  # Required Python libraries
 ├── config.toml       # Streamlit config
 ├── README.md         # Project documentation
 ├── LICENSE           # License file

🧠 Technologies Used

• Python 🐍
• Streamlit – UI for interaction
• PyPDF2 – PDF text extraction
• NLTK – Sentence segmentation
• Sentence Transformers – all-mpnet-base-v2 for semantic similarity
• Hugging Face Transformers – google/flan-t5-large for generative responses

🌱 Future Improvements

✅ Add support for scanned PDFs via OCR
✅ Expand to handle multiple documents
✅ Let users select different AI models
✅ Deploy online via Hugging Face Spaces or Streamlit Cloud

🙋‍♂️ About the Creator

📌 Shantanu Mohod – Passionate about AI, Data Science, and Full-Stack Development
🔗 GitHub: shantanumohod23
🔗 LinkedIn: Shantanu Mohod