Project Overview

AI-Powered Assistant for Akdeniz University CSE Students

CSE-351 Design Patterns Term Project

ChatbotForCse is an AI-powered assistant designed for Akdeniz University Computer Science students. This project demonstrates the practical application of three core design patterns.

Strategy Pattern

Dynamic algorithm selection for intent classification and context fetching

Factory Pattern

OCP-compliant object creation with dictionary-based registry

Singleton Pattern

Thread-safe resource management for database and LLM clients

Layered Architecture

Clean separation of API, Service, Domain, and Data layers

---

Problem Statement

The Challenge

Students at Akdeniz University struggle to access scattered information:

Problem	Current Solution	Pain Point
Daily dining menus	Check SKS website	Manual, time-consuming
Department announcements	Monitor CSE website	Easy to miss
Academic questions	Ask in WhatsApp groups	Repetitive, slow

Why This Project?

Existing solutions require:

• Checking multiple websites daily
• Asking repetitive questions in group chats
• Waiting for human responses

Our solution: A unified AI chatbot that aggregates all information sources.

---

Solution: ChatbotForCse

A unified AI-powered chatbot accessible via WhatsApp and Mobile App.

Features

Dining Menus

Scrapes weekly menus from SKS, parses images using Gemini Vision

Announcements

Tracks and delivers department news in real-time

Intelligent Q&A

Uses Gemini and Ollama for general questions

Multi-Channel

Live on WhatsApp, API ready for mobile

Tech Stack

Component	Technology
Backend	FastAPI (Python)
Database	MongoDB
AI Chat	Google Gemini 2.5 Flash
AI Classification	Ollama (Qwen2 7B)
WhatsApp	WPPConnect (Node.js)

Architecture

---

Design Pattern Summary

Each pattern solves a specific problem in our architecture. Click the cards above to learn more.

Pattern	Category	Problem Solved	Benefit
Strategy	Behavioral	Multiple classification algorithms	Runtime flexibility
Factory	Creational	Complex object creation	OCP compliance
Singleton	Creational	Expensive resource initialization	Thread safety

---

Quick Start

1. Clone Repository

   git clone https://github.com/Akdeniz-CSE-Students/csebot_documentation/.git
   cd ChatbotForCse
   
2. Start Backend

   cd backend
   pip install -r requirements.txt
   uvicorn main:app --reload
   
3. Start WhatsApp Client

   cd whatsapp
   npm install && npm start
   

See Implementation for detailed setup instructions.