TeachHubl, 05/2025-07/2025, Built TeachHubl, a web platform for one-on-one virtual home tuition focusing on delivering real-time and personalized learning experiences. Implemented WebRTC for seamless video and audio calling, and WebSockets to enable live chat and an interactive whiteboard for collaborative teaching. Developed a scalable Node.js and MongoDB backend supporting custom test creation and student performance tracking. Designed a React frontend to visualize dashboards and enable smooth class interactions. Integrated email-password and Google OAuth authentication using JWTs and sessions for secure access. Built a custom push notification system with service workers for real-time alerts, and integrated AWS SES, Amazon S3, and Razorpay to manage transactional emails, file storage, and in-app payments securely. Crafts’ n stitch, 03/2024, Crafts' n Stitch is a web platform that enables users to design custom outfits or personalize garments created by designers, combining creative freedom with e-commerce functionality. The React frontend delivers a responsive and user-friendly interface for browsing, customizing, and previewing clothing designs. The backend, built with Node.js, ensures scalable server-side operations, while Sequelize ORM with MySQL manages structured data such as user profiles, design templates, and orders efficiently. Razorpay was integrated to handle secure, real-time payments between customers and designers. Brain Tumour Prediction and Cancer Stage Classification, 09/2024, Developed a machine learning model for brain tumour symptom identification and cancer stage prediction, focusing on gliomas. Engineered features and applied SHAP values for model interpretability and explainability. Pre-processed MRI images using median filtering and wavelet transform to enhance model performance.