Built a machine learning-based recommendation system for personalized anime suggestions.
Utilized Scikit-learn, Pandas, Django, and Streamlit to analyze 5M+ user ratings and suggest relevant anime.
Implemented a hybrid approach using collaborative filtering and content-based similarity.
Developed a web application to provide optimal insurance recommendations for uninsured patients.
Utilized OpenAI models, Next.js, and FastAPI to streamline decision-making for medical professionals.
Winner of UB AI Hackathon
Created an interactive web app for handwritten doodle recognition.
Built with a Resnet18 model trained in PyTorch with MNIST dataset, integrated with a user-friendly front-end.
Hosted on Huggingface spaces.
Developed a React + TypeScript application to visualize sorting algorithms.
Used Chart.js for real-time graph updates and deployed the app using GitHub Pages.
Helps users understand sorting concepts interactively.