Intelligent Real-Time Resource Management in Cloud Computing: AI-Driven Optimization for EC2, RDS, and ECS
This project focuses on optimizing AWS cloud resources (EC2, RDS, and ECS) using AI-driven techniques. It addresses inefficiencies in traditional cloud resource management by leveraging LSTM models for workload forecasting and RL for dynamic resource scaling. The system integrates AWS CloudWatch for data collection, processes real-time metrics, and automates scaling to enhance performance, cost efficiency. This project focuses on optimizing AWS cloud resources using AI-driven techniques.
DriveWise
DriveWise: Smart Vehicle Maintenance Application
DriveWise is an innovative smart vehicle maintenance application designed to simplify and enhance the vehicle upkeep experience for everyday users. The core idea of DriveWise is to create a seamless connection between vehicle diagnostics, users, and trusted service providers using a technology-driven approach. The application integrates with vehicle sensors to monitor performance and health in real-time, decoding trouble codes (like OBD-II) to detect potential issues before they escalate.
One of DriveWise’s key features is its intelligent alert system, which notifies users immediately when a problem is detected, along with a clear explanation and potential causes. To reduce stress and time spent on repairs, the app automatically matches users with nearby, verified service providers who specialize in the identified issue. Users can view repair shop ratings, estimated costs, and even schedule appointments directly from the app.
DriveWise also keeps a digital maintenance history, helping users track past repairs, routine services, and upcoming maintenance schedules. By leveraging data analytics and predictive maintenance, the app aims to reduce unexpected breakdowns and extend vehicle lifespan.
The project is developed by a team of six undergraduate students and focuses on providing a user-friendly, efficient, and reliable solution for vehicle maintenance. DriveWise stands out for its combination of real-time diagnostics, smart recommendations, and integration with local mechanics — making car maintenance smarter, faster, and stress-free.
BlockBallot
BlockBallot is a next-generation blockchain-based electronic voting system developed to revolutionize Sri Lanka’s electoral process. Traditional elections in Sri Lanka rely heavily on manual, paper-based procedures, resulting in high costs, significant logistical challenges, security vulnerabilities, and delays in result announcements. BlockBallot directly addresses these issues by harnessing the power of blockchain technology, QR-based voter verification, and real-time digital vote counting to ensure a transparent, secure, and efficient voting experience for all citizens.
The system enables each registered voter to be authenticated through a unique QR code, eliminating manual errors and reducing the potential for fraud or duplicate voting. Votes are cast electronically on tablets at polling stations, instantly encrypted and immutably stored on the Ethereum blockchain. This automated process delivers real-time vote counting, immediate results, and comprehensive audit trails—removing the delays and inaccuracies of manual counting. Voters receive a printed receipt as confirmation, enhancing trust and transparency.
Accessibility and inclusivity are central to BlockBallot’s design, featuring a multilingual interface (Sinhala, Tamil, and English) and alternative voter identification options to accommodate all demographics, including those without traditional ID cards. The solution’s robust security measures, such as end-to-end encryption and role-based access controls, ensure compliance with both local and international data protection standards.
By digitizing the voting process, BlockBallot significantly reduces costs, environmental impact, and barriers to participation, while enabling rapid, reliable, and tamper-proof elections. Ultimately, BlockBallot empowers every Sri Lankan to participate confidently in a secure, transparent, and modern democratic process.
Retail Strategy Optimization via ML Customer Segmentation
In the evolving retail landscape, businesses must adapt to changing consumer behavior to remain competitive. Our project, Optimizing Retail Strategy through Machine Learning-Based Customer Segmentation, focuses on empowering John Keells, a leading supermarket chain in Sri Lanka, to enhance its marketing effectiveness through data-driven insights. With operations across 20 locations and a wide product range—including dry goods, fresh produce, and luxury items—John Keells faces the challenge of addressing diverse customer needs using traditional, generalized marketing strategies.
To overcome this, we developed a machine learning-based solution that segments customers according to their purchasing behavior using historical sales data. The process involved comprehensive data preprocessing, including cleaning, feature engineering, and normalization, followed by model development. We evaluated multiple classification algorithms, including Random Forest, XGBoost, LightGBM, CatBoost, and Neural Networks.
These insights enable John Keells to adopt highly targeted and personalized marketing campaigns, improve customer retention, and optimize inventory planning. The solution also provides a foundation for future CRM and marketing automation efforts. This project showcases how machine learning can transform traditional retail strategies into intelligent, customer-focused approaches that drive business growth.
UniGuide
UniGuide is a student-focused platform that helps individuals make smart educational and career decisions. It offers a comprehensive database of universities, real-time updates on courses and trends, and clear learning paths aligned with career goals. Designed to simplify complex choices, UniGuide brings everything students need—from degree options to student insights—into one easy-to-use platform.
AuraMirror
Aura Mirror is a smart mirror project designed to seamlessly integrate technology into everyday routines, offering a personalized and informative user experience. Developed using a Raspberry Pi 4, a display screen, and a two-way mirror, it functions as both a conventional mirror and a digital display for essential daily information such as time, date, weather updates, calendar events, news headlines, and to-do lists.
The mirror supports facial recognition to deliver tailored content for each user. It also includes entertainment features such as music and video playback. A mobile application developed in Kotlin listens to notifications from the user’s smartphone and forwards them to the mirror. The backend is developed using Spring Boot, which handles data processing and communication between the mobile app and the mirror.
The front end is built using React and React Native, ensuring a responsive and user-friendly interface. Aura Mirror combines IoT components, modern web technologies, and mobile integration to create a functional and visually appealing smart device that enhances everyday living with convenience and personalization.
AidPal
AidPal is a smart, AI-powered first aid assistant designed to guide users through emergency situations by providing instant, accurate, and easy-to-follow medical guidance. Whether someone is dealing with a burn, cut, bruise, sprain, or other common injury, AidPal empowers users to take the right steps before professional help arrives.
Users can simply upload or take a photo of the injury, and AidPal uses advanced image recognition to identify the type of wound. Once detected, the app delivers step-by-step first aid instructions based on globally trusted medical guidelines, including those from the Red Cross and World Health Organization.
If image recognition is unclear or unavailable, users can type in their symptoms or use the built-in chatbot to receive relevant help. The app also offers voice-guided instructions and, where applicable, 3D visual aids to make first aid even easier to follow.
AidPal also includes an emergency mode that enables users to quickly access and display instructions for critical situations. Designed for speed, clarity, and accessibility, AidPal transforms any phone or computer into a life-saving companion, bringing essential first aid knowledge to users anytime, anywhere.
ThoraxAI
ThoraxAI is an advanced artificial intelligence-powered system designed for the automated detection and diagnosis of thoracic diseases using chest X-ray images. Developed with the goal of addressing the global shortage of radiologists and the rising demand for rapid and accurate diagnostics, ThoraxAI leverages deep learning using convolutional neural networks (CNNs) to analyze medical imaging data with high precision.
The system is trained on two large-scale, annotated chest X-ray datasets, namely NIH ChestX-ray14 and CheXpert, enabling it to recognize a wide spectrum of thoracic abnormalities, including pneumonia, tuberculosis, lung nodules, cardiomegaly, and pleural effusion. Through image preprocessing, feature extraction, and multi-label classification, ThoraxAI provides interpretable diagnostic outputs, including heatmaps and confidence scores, to assist clinicians in decision-making.
ThoraxAI is designed for integration into hospital radiology workflows, allowing real-time triage of urgent cases, reduction in diagnostic errors, and acceleration of reporting times. Its modular architecture supports both on-premise deployment in medical institutions and scalable cloud-based solutions for remote or underserved regions. Beyond diagnosis, ThoraxAI incorporates explainable AI components to enhance trust and usability among healthcare professionals by offering visual explanations of its predictions.
By combining technological innovation with practical clinical utility, ThoraxAI addresses key challenges in global healthcare, especially in low-resource settings by offering a scalable, efficient, and cost-effective diagnostic aid that empowers radiologists, reduces workload, and ultimately improves patient outcomes.
Trip Ceylon
Trip Ceylon is an intelligent travel assistant platform designed to transform how travelers experience Sri Lanka. Whether you’re planning your dream vacation or already exploring the island, Trip Ceylon empowers you with tools for seamless, secure, and socially connected travel.
At its core, Trip Ceylon offers an AI-powered travel plan generator that creates fully personalized itineraries based on your preferences, budget, desired pace, and travel duration. From golden beaches to misty hill country, our system curates the ideal route with must-visit spots and hidden gems.
The platform also includes a smart luggage tracking system, utilizing GPS-enabled tags to monitor your baggage in real time—helping ensure peace of mind throughout your journey. Whether hopping between cities or arriving at the airport, you’ll always know where your bags are.
Trip Ceylon is more than just a tool—it’s a community platform. Users can connect with nearby travelers, form new friendships, and even join local volunteering projects. The app also serves as a sharing space where travelers can post their experiences, reviews, and tips or follow others’ travel diaries to get inspired.
Built for travelers by travelers, Trip Ceylon brings together safety, personalization, and social interaction—making it the ultimate companion for anyone visiting Sri Lanka.
Spark Voyage
Spark Voyage is an interactive, AI-powered educational platform tailored to support preschool children (ages 3–5) with Attention-Deficit/Hyperactivity Disorder (ADHD). The solution addresses attention-related challenges through gamified learning, personalized guidance, and real-time insights for caregivers. At its core is an immersive mobile adventure game where children are guided by “Spark,” an AI-driven genie who adapts instructions, encouragement, and emotional feedback based on the child’s interaction and focus levels.
Designed to foster cognitive skills such as attention, memory, and early literacy, the platform uses storytelling, puzzles, and adaptive mini-games to maintain engagement and promote meaningful learning outcomes. These activities are grounded in child development research and incorporate culturally relevant themes to ensure accessibility, particularly in the Sri Lankan context.
An integrated parent dashboard provides visual progress reports, helping caregivers track cognitive and emotional development and understand engagement patterns. By utilizing Internet of Things (IoT) principles, Spark Voyage captures behavioral data from the learning environment to refine personalization and enhance child-caregiver communication.