My Portfolio
A showcase of my recent work spanning web applications, mobile apps, and enterprise solutions. Each project represents a unique challenge solved with modern technologies and best practices.
FlySpotter Pro
FlySpotter Pro revolutionizes fly identification using TensorFlow Lite-powered machine learning for instant, accurate results. Built with Flutter and Firebase, it offers offline capability, cross-platform support, user management, discovery tracking with Google Maps integration, and a rich educational guide. Designed for entomologists, researchers, pest control professionals, and nature enthusiasts.
Technologies Used
CUT Portal WhatsApp Bot
CUT Portal WhatsApp Bot enables students to navigate their portal features — including profiles, course details, grades, and announcements — via WhatsApp. Built with a clean architecture approach, the system separates concerns into models, services, controllers, and routes. Designed for scalability and real-time support using the WhatsApp Business API.
Technologies Used
Smart Driver Companion
Smart Driver Companion is a comprehensive Flutter application that integrates AI-driven driver behavior monitoring, real-time community road alerts, turn-by-turn navigation, and a service locator for drivers. Featuring GetX for state management and a clean MVC architecture, the app leverages Firebase, Google Maps APIs, and a custom Python-based behavior scoring API to deliver an engaging, safety-focused driving experience.
Technologies Used
ZINWA Water Meter System
The ZINWA Water Meter System is a comprehensive platform combining an Express.js backend API with a Flutter mobile app. It empowers customers to manage their water accounts, purchase prepaid tokens, monitor usage, and receive alerts. The backend handles authentication, payments via Paynow, real-time meter reading, analytics, and administrative functions, while the mobile app offers an intuitive user experience with offline support, biometric login, and multi-language support.
Technologies Used
LearnSmart AI-Powered e-Learning Platform
LearnSmart is a full-stack monorepo education platform with AI-powered features for personalized learning and management. The system features separate Next.js frontends for administrators and lecturers, powered by Redux, Radix UI, Tailwind, and more. The backend is built with Node.js, Express, and Sequelize ORM, featuring robust security, OpenAI integration, JWT-based authentication, Supabase for storage, and Paynow for payment processing. Designed for scalability, the platform supports role-based access, learning management, AI integration for enhanced learning experiences, and a modern, maintainable UI/UX.
Technologies Used
Energy Monitor System
Energy Monitor is a multi-platform energy tracking solution aimed at promoting sustainability and efficient energy use. The system includes a Flask-based web application for visualizing energy usage trends and analytics, and a Flutter-based mobile application for monitoring on the go. It provides real-time tracking, historical analysis, device-level monitoring, alerts, and intelligent recommendations. The backend supports secure authentication, RESTful APIs, and integration with smart devices. Built with scalability in mind, Energy Monitor empowers homeowners, facility managers, and organizations to make data-driven decisions on energy consumption.
Technologies Used
University Student Portal Web Analytics System
This system provides comprehensive real-time and historical analytics for a university student portal. It tracks active users, clicks, session durations, and navigation flows through a browser extension that streams events via WebSockets. The React dashboard visualizes these insights with charts and heatmaps. The backend, powered by Express and Supabase (PostgreSQL), manages data storage, authentication, and real-time communication. The architecture is optimized for scalability and secure data handling, offering role-based access to administrators and authorized personnel.
Technologies Used
tflite Flutter Plugin
The tflite Flutter plugin allows Flutter apps to perform on-device ML inference using TensorFlow Lite models. It supports a variety of vision tasks such as image classification, object detection using SSD MobileNet and Tiny YOLOv2, image-to-image translation with Pix2Pix, semantic segmentation with Deeplab, and human pose estimation via PoseNet. The plugin is compatible with iOS and Android and includes support for GPU acceleration and real-time camera streams.
Technologies Used
GeoFlutterFire3
GeoFlutterFire3 is a Flutter library that simplifies storing and querying geospatial data in Firebase Firestore. It enables real-time geolocation querying with automatic updates, making it suitable for location-based apps such as delivery tracking, ride-hailing, and social platforms. The library extends Firestore functionality without altering existing schemas or security rules and provides efficient querying with geohash-based lookups. Inspired by GeoFireX, it offers both reactive streams and single subscriptions with improved memory safety.
Technologies Used
Lucid Eye — AI Assistance App for the Blind
Lucid Eye is an AI-powered mobile assistant aimed at helping visually impaired users interact with their surroundings. Initially built as a university project, the app evolved into a multi-functional tool offering AI-based object detection, text recognition, currency detection, maps navigation, chatbot communication, and an SOS feature. It provides real-time assistance, leveraging machine learning and external APIs to create a meaningful impact for the blind community.
Technologies Used
LocalGenAI — Offline AI Assistant
LocalGenAI is a privacy-first AI assistant enabling users to chat with multiple AI language models without requiring an internet connection. Built with Flutter and leveraging mobile-optimized AI models (via ONNX, TensorFlow Lite, Core ML), it offers a fully offline experience. Users can download models, switch between them, and enjoy seamless interaction on both Android and iOS. The app ensures all data remains on-device, making it ideal for users concerned about privacy.
Technologies Used
St Joseph's Mission Hospital Website
A clean, fast, and mobile-friendly website designed for St Joseph's Mission Hospital. Built using React with TypeScript and powered by Vite for optimal development and build speed. The site provides essential information about hospital services, departments, contact details, and general health awareness content.
Technologies Used
My Development Process
Discovery
Understanding your needs, goals, and target audience
Planning
Creating wireframes, architecture, and project timeline
Development
Building your application with regular updates and feedback
Launch
Deployment, testing, and ongoing support