All Projects
BitPort is a modern, full-featured digital marketplace where creators can sell and buyers can purchase digital products like PDFs, templates, code files, graphics, and more. Built with React, TypeScript, and powered by Supabase with secure Stripe payment integration, it provides a seamless experience for both sellers and buyers.
AI-powered tool for generating slide decks from a single prompt using Gemini. Offers advanced editing and export to PNG, PDF, and PPTX.
A web application built with Streamlit to demonstrate Optical Character Recognition (OCR) capabilities using the Tesseract engine. Users can upload images, preprocess them, and extract text with bounding box visualization.
A Python-based SMTP server designed for secure, efficient, and user-friendly email communication with optimized network handling and basic authentication.
A simplified operating system developed in C that supports basic command execution, memory management, and process scheduling.
A basic version of Excel built in C++ using linked list structures to manage cell data, perform arithmetic operations, and mimic spreadsheet functionality.
A minimal MS-DOS/CMD clone developed in C++ that supports basic shell commands, file operations, and directory management.
AI-powered COVID-19 detection system using chest X-rays and deep learning models.
Personalized chatbot for elderly care using natural language processing and machine learning.
Data science project to predict HIV incidence rates using demographic and healthcare data.
A desktop-based system for managing car sales data, generating sales reports and trends.
A C#-based desktop application with database integration to manage Course Learning Outcomes (CLOs), map them with Program Learning Outcomes (PLOs), and record assessment results for academic performance evaluation.
Adobe Illustrator plugin for WinPlot integration to manage file output with registration mark support.
Tailored the open-source Kinovea motion analysis software to meet a client's unique sports performance tracking needs. Implemented custom measurement tools, improved user interface components, and optimized video processing for smoother playback and annotation.
A database-driven system designed to detect and prevent fraudulent activities in both digital and traditional financial transactions. Utilizes a rule-based engine to flag suspicious patterns and provides analysts with a real-time dashboard for monitoring threats.
A Python Flask-based web application that classifies handwritten digits from the MNIST dataset. The system is secured using Zero Trust Architecture principles, including strict identity verification, microsegmentation, and minimal trust boundaries to prevent unauthorized access.