General Projects

A web app where users can upload documents and get context-aware text mentoring. Built with Next.js for the frontend and powered by a RAG backend, it retrieves relevant document sections to generate intelligent suggestions.

A smart home control system where ESP32 devices manage sensors and appliances, connected to a Firebase database for real-time updates. The React.js web app provides users with a dashboard to monitor and control devices remotely.

A Retrieval-Augmented Generation (RAG) application designed for a travel company to deliver personalized recommendations and answer customer queries.

A gas detection system built using an operational amplifier (Op-Amp), simulated in Tinkercad. The system detects varying gas concentration levels and triggers alerts (LED or buzzer) based on predefined thresholds for safety monitoring.

A Retrieval-Augmented Generation (RAG) system that enhances medical document analysis by combining vector-based search with a large language model. It retrieves relevant medical records.

A dashboard application built with Django as the backend and React.js as the frontend to visualize and manage given datasets. It provides interactive charts, filters, and reports, enabling users to gain insights and track key metrics in real time.

An automated web scraping workflow built with n8n, enabling data extraction from websites. The workflow can parse, transform, and store scraped data into google sheet.

A Python-based scraper that uses Selenium to handle dynamic, JavaScript-rendered pages and BeautifulSoup to parse the HTML and extract structured data efficiently.

A full-stack company management system built with Django for the backend and Next.js for the frontend. It streamlines operations such as employee management, project tracking, and reporting, with secure authentication and a modern user interface.

A proxy-backed web scraper that uses Selenium to load dynamic web pages and BeautifulSoup to parse and extract data.

A web-based ticket booking system built with Django, providing users with a simple interface to browse events, book tickets, and manage reservations.

A desktop-based supermarket management system built using Python Tkinter, designed to handle product listings, billing, and inventory tracking.

A RAG-powered application for restaurants that uses n8n to automate workflows, retrieve menus, reviews, and policies, and generate personalized, context-aware responses to customer queries.

A face detection system built with Python, leveraging OpenCV and MediaPipe to accurately detect and track human faces in real time.

A product query application built with n8n that connects Telegram for customer queries, fetches product details from Google Sheets, and sends automated responses or follow-ups through Gmail.

A machine learning project that performs text classification to analyze sentiment (positive, negative, or neutral).

A car parking assistant system built with Arduino and simulated in Proteus. It uses ultrasonic sensors to detect obstacles and measure distance, providing alerts to assist safe parking.

A voice recognition system developed in Python using neural networks to process and classify audio inputs. The model extracts speech features, trains on labeled datasets, and accurately identifies or verifies spoken commands.

Developed an AI Personal Assistant using Next.js and the OpenAI Agent SDK that can manage Google Calendar appointments, analyze Excel datasets, generate data visualizations and insights, and create or update notes in Google Documents. The agent uses natural-language interaction to understand user requests and execute multi-step tasks across different services.

Developed an IoT-based Smart Fridge system that monitors food items and detects their expiration status using a trained machine learning model. The system uses Raspberry Pi for local hardware control and Firebase Realtime Database for storing and synchronizing data. A React.js web application provides real-time monitoring and displays alerts when food items are expired, helping users track food freshness and reduce waste.

Developed an AI-powered epilepsy app recommendation system using Python, OpenAI Agent SDK, and Gradio. The system uses multiple AI agents in a sequential pipeline to conduct a patient interview, identify required features, search for relevant epilepsy applications, and rank them based on their compatibility with the user’s needs. Pydantic is used for structured data validation, while Python asyncio enables asynchronous processing to improve response time. The system processes patient information only during the active session without permanently storing personal data.

Developed a full-stack tourism management web application using a three-tier architecture with Next.js, TypeScript, Tailwind CSS, Django REST Framework, and PostgreSQL. The system provides bilingual Arabic/English support with RTL, JWT authentication, role-based access control, and full CRUD functionality for managing tourist attractions, categories, reviews, and favorites. Google Maps API is integrated to display attraction locations and navigation, while Docker, Nginx, and HTTPS are used for secure and scalable deployment on a Linux VPS.

Developed a low-cost IoT-based smart home system using an ESP32 to monitor environmental conditions such as temperature, humidity, smoke, motion, and light intensity, while controlling devices including a fan, lights, buzzer, and servo motor. A real-time web application was developed using React.js and Tailwind CSS, with Firebase Realtime Database enabling continuous data synchronization, remote device control, and real-time monitoring from anywhere.

Developed an IoT-based Smart Waste Monitoring System for real-time monitoring of waste bins and environmental conditions to improve waste collection efficiency. The system uses ESP32 with an HC-SR04 ultrasonic sensor to detect bin fill levels, and MQ-2 and MQ-135 sensors to monitor smoke and gas concentrations. Sensor data is transmitted in real time to Firebase Realtime Database, while a React.js web application provides an interactive dashboard and map for monitoring bin status, environmental alerts, bin locations, and waste collection routes.

Developed an AI-powered Transformer Fault Diagnosis Agent using Python and machine learning to classify transformer conditions from five Dissolved Gas Analysis (DGA) measurements (H₂, CH₄, C₂H₆, C₂H₄, and C₂H₂) into seven fault categories aligned with IEC 60599. The system uses an enhanced Random Forest model with SMOTE and gas-ratio, log-transformed, and interaction features, achieving 86.32% classification accuracy. The best model was deployed through a Gradio web interface, with GPT-4o-mini generating plain-language diagnostic reports from the model predictions, making the system accessible to both engineers and non-expert users.

This project demonstrates an Arduino-based smart pump control and protection system. The operator can start and stop the pump using push buttons. After the pump is stopped, the controller waits 5 seconds before activating a separate test relay for cable short-circuit monitoring. A DIP switch is used to simulate a cable short circuit. If a short is detected, the system stops the pump, blocks restarting while the fault is active, activates a red LED alarm, and displays the system status on the Serial Monitor. The prototype demonstrates pump protection, cable fault detection, automatic interlocking, and alarm indication.