Çimsa Smart Assistant 🤖💻

This project is a desktop chatbot application developed to provide quick solutions for common IT support issues at Çimsa.
It combines rule-based logic with a lightweight NLU (Natural Language Understanding) approach.
The application also features a modern CustomTkinter interface, offering a user-friendly chat experience. 🎯

---

🚀 Features

• Quick answers to common IT issues:
  • VPN connection problems
  • Outlook password reset
  • SAP login issues
  • Printer / OneDrive / Teams problems
• Ticket Creation Process:
  • Step-by-step collection of Title → Detail → Priority (urgent/normal)
  • Optional screenshot attachment
• NLU Support:
  • Can handle typos (e.g., vpn bağlnmıyo)
  • Provides clickable suggestions when unclear
• Corporate Information Module:
  • Çimsa mission & vision
  • Website
  • Company history
  • CEO information
  • Sabancı Holding background
• Modern UI Design:
  • CustomTkinter-based, dark theme support
  • User & bot message bubbles
  • Quick access buttons in sidebar (VPN, Outlook, SAP, etc.)

---

🛠️ Technologies Used

• Python 3.11
• CustomTkinter → modern GUI
• Pillow (PIL) → logo & image handling
• fuzzywuzzy + python-Levenshtein → fuzzy text matching
• Regex (re) → rule-based intent routing
• JSON → intent & response database

---

📂 Project Structure Cimsa_Chatbot/ │ ├── gui_app.py # GUI (CustomTkinter) ├── chatbot_logic.py # Chatbot logic (intents + NLU + rules) ├── faq.json # Intent-response dataset ├── requirements.txt # Required libraries ├── logo.png # App logo └── talep_log.txt # Ticket logs (optional)

---

⚡ Installation Clone the repository:

git clone https://github.com/username/cimsa-chatbot.git
cd cimsa-chatbot


Create and activate a virtual environment:
python -m venv venv
source venv/bin/activate  

Install requirements:
pip install -r requirements.txt

Run the application:
python gui_app.py

---
🔮 Future Improvements: 
Web-based version (Flask / FastAPI + React UI)
Integration with Microsoft Teams / Slack
Advanced NLU (spaCy / scikit-learn / HuggingFace)



---
⚠️ Limitations
This chatbot was developed within a limited internship period.  
Due to time constraints, it may not always provide 100% correct answers.  
The main goal is to demonstrate the integration of rule-based logic, NLU, and a modern GUI in a practical IT support scenario.

---
👤 Author: Kağan Kılıç
🎓 Computer Engineering, Eskişehir Technical University

🔗 LinkedIn: linkedin.com/in/kagankilic

💻 GitHub: github.com/klckagan

📜 License

This project was developed for educational and internship purposes.