Fitness & Nutrition Platform
Multi-role web application featuring personalized workout & nutrition plans, subscription management with Stripe & bKash payment integration, BMI tracking, video tutorials, discussion forum, and AI chatbot assistance.
Computer Science Engineer | Securing Digital Frontiers | Quality-Driven Development
I graduated with a B.Sc in Computer Science and Engineering from BRAC University (2021-2026), earning a CGPA of 3.60. Over the course of my degree, I gained hands-on experience across full-stack web development, machine learning and deep learning, and software quality assurance, building real products end to end rather than just studying them in theory.
I care deeply about writing secure, maintainable code and understanding systems from multiple angles — whether that means architecting a multi-role web platform, designing and training intelligent models, or stress-testing an application for vulnerabilities.
Multi-role web application featuring personalized workout & nutrition plans, subscription management with Stripe & bKash payment integration, BMI tracking, video tutorials, discussion forum, and AI chatbot assistance.
Multi-role emergency response web application with real-time helper tracking, AI-powered medical chatbot for first aid guidance, hospital locator with directions, and premium feature payment interface integration.
Comprehensive rental property management platform with secure user authentication, detailed property listings, advanced search filters, booking request system, and admin dashboard for property management.
View Project →High-precision customer churn prediction system achieving ~98% F1-score using cost-sensitive Random Forest models, advanced feature engineering, and interactive real-time Streamlit dashboard for business insights.
View Project →Role-based platform for educational institutions covering the full assignment lifecycle — admins manage users, classes and teacher mappings, teachers create and grade assignments with deadline and late-policy controls, and students submit work and track grades, all secured with JWT role-based authorization.
Real-time road safety solution using YOLOv8 segmentation for computer vision-based pothole detection, tracking, and risk classification with automated severity reporting and analytics dashboard.
View Project →IoT-based parking management system using Arduino and RFID technology for contactless vehicle entry/exit, automated toll collection, and real-time parking slot availability monitoring.
Interactive web application for exploring comprehensive global country information including demographics, geography, economy, and cultural insights through an intuitive and responsive interface.
Classic tile-based puzzle game built with Python and Pygame featuring intuitive controls, time-based challenge, score tracking system, and smooth animations for an engaging gaming experience.
Automated news aggregation and headline integrity monitor that scrapes six major sources with Playwright, delivers clean article cards to Telegram, and flags publisher headline drift using fuzzy string matching against live RSS feeds, fully automated via a scheduled GitHub Actions pipeline.
View Project →Monorepo combining a Playwright UI regression suite and a Postman/Newman API test suite for a Django backend, wired into dual GitHub Actions pipelines that run across a Chromium/Firefox/WebKit browser matrix and upload test reports automatically on every push, pull request, and nightly run.
View Project →Complete QA strategy artifact built around the Fitness & Nutrition Platform, featuring risk-based test planning, a requirements traceability matrix, manual test cases, an independently re-verified bug report, Playwright regression automation, k6 performance testing, and an axe-core accessibility audit.
View Project →Developed a continuous authentication system leveraging deep learning on behavioral biometric signals - specifically keystroke dynamics and touchscreen interaction patterns. For keystroke dynamics, a hybrid Siamese BiLSTM + CNN architecture was applied, while touch gestures were modeled using a Siamese LSTM. The model achieves strong cross-session and unseen-user generalization, demonstrating the feasibility of passive, non-intrusive user authentication.
Dhaka, Bangladesh
+880 1407 963 892
Scan this QR code to visit GitHub profile
github.com/FaisalAhmed21