Tetaş AI Chatbot 🤖

Tetaş AI Chatbot is an intelligent assistant designed to help users easily access product information from a set of Turkish PDF catalogs. Using OCR, chunking, vector embeddings, and OpenAI's GPT model, it enables natural language interaction with technical product data.

🚀 Features

• ✏️ Extracts text from multiple Turkish PDF catalogs using OCR
• 📂 Splits text into meaningful chunks for accurate retrieval
• 🧠 Embeds chunks into a FAISS vector store for semantic search
• 🤖 Uses GPT-3.5-turbo to answer questions based on relevant document chunks
• 🛍️ Provides a Streamlit-based web interface for interactive Q&A

🧠 Technologies Used

• Python 3.11
• LangChain
• OpenAI (GPT-3.5 Turbo)
• FAISS (Facebook AI Similarity Search)
• PyMuPDF + pytesseract for OCR
• Streamlit for the web interface
• dotenv for secure API key management

⚙️ Setup & Installation

1. Clone the Repository

git clone https://github.com/BartugKaan/tetasChatbot.git
cd tetasChatBot

2. Install Requirements

pip install -r requirements.txt

If you don’t have FAISS installed, you may need:

pip install faiss-cpu  # or faiss-gpu if supported

3. Add Your OpenAI API Key Create a .env file in the root directory:

OPENAI_API_KEY=your_openai_key_here

⚡ How to Use

Step 1: Process the PDFs (one-time)

python main.py

This step will extract text, chunk it, embed it, and save the FAISS vector store.

Step 2: Run the Chatbot Web App

streamlit run app.py

Go to http://localhost:8501 and start asking questions!

🙌 Credits

This project was developed as part of an AI engineering practice to explore Retrieval-Augmented Generation (RAG) pipelines using LangChain, OpenAI, and PDF data. Also special thanks for Tetaş for PDF Catalogs.

📍 License

MIT License