React, Express, Node.js, MongoDB, JavaScript, Clerk, LLM
A web app that helps users plan meals and manage groceries based on dietary preferences, budget, and cooking habits. It uses AI to generate shopping lists, suggest restocking, and reduce food waste through personalized recommendations.
AWS, Node.js, CI/CD, MySQL, Docker
A system level transformation of a legacy app into modular, cloud native microservices using AWS. It includes containerized deployments, automated CI/CD pipelines, and load-balanced, zero downtime updates via ECS Fargate and ALB.
Python, StreamLit, NLP, HuggingFace, FAISS
A PDF-based chatbot that lets users ask questions about document content. It processes and embeds PDFs using language models, retrieves relevant sections with vector search, and responds with context-aware answers using Flan-T5-XXL.
MongoDB, Express.js, React, Node.js, Clerk Auth
A full-stack finance tracker app that lets users securely log in, add income/expenses, and visualize their financial activity. Built with Clerk for seamless authentication, the app includes persistent storage, dynamic balance updates, and responsive UI.
Python, NLP, NER, Pandas, NumPy
An NLP pipeline that extracts and analyzes symptom data from vaccine adverse event reports. It uses biomedical named entity recognition and word embeddings to identify patterns across demographics and evaluate mapping accuracy.
Python, K Means, Pandas
A data science project that creates new performance metrics for NBA players using K-means clustering and game context (like shot location and defender proximity). Results were benchmarked against ESPN rankings using Rank Biased Overlap.
Java, Android, Firebase
A real-time Android note-taking app built with Firebase for storage and authentication. It supports secure user logins, media-embedded notes, and a responsive UI with RecyclerView for managing and syncing notes across devices.
Python, CNN, LSTM, Apps Script
A deep learning system that recognizes handwritten mathematical expressions using a CNN-LSTM model. It was trained on over 137,000 auto-generated samples and reached over 90% validation accuracy through advanced preprocessing and tuning.
A chatbot that analyzes user sentiment during conversations and suggests music tailored to their mood by integrating with external music APIs.
A web-based version of Flappy Bird built with JavaScript and Canvas API, where players navigate a bird through obstacles by controlling its flight with key presses.
An interactive web app for managing tasks, allowing users to add, mark complete, and delete to-dos with real-time updates and intuitive UI.
A predictive web app that uses a Random Forest model to assess diabetes risk based on user health inputs, providing an easy-to-use interface for health screening.
A content-based movie recommender that suggests films similar to a user’s choice by comparing movie features like genres, cast, and keywords using cosine similarity.
An analysis and prediction tool for COVID-19 case trends in India, leveraging data visualization and machine learning to forecast future outbreaks.
A responsive quiz app that dynamically presents GRE practice questions, tracks progress, and provides immediate feedback to aid exam preparation.
A recommendation system that suggests accommodations by grouping similar listings through clustering and integrating geolocation data to match user preferences.
A React app that lets users search recipes via the Edamam API, displaying detailed results including ingredients and images in a user-friendly interface.
A project analyzing YouTube comments to detect sentiments and discover discussion topics using NLP techniques and clustering algorithms.
A general store web system supporting user registration, secure login, session management, and employee data handling using PHP and MySQL.
A web app that allows users to create, preview, and print customized resumes with rich text editing features and dynamic form inputs.
MATLAB, Computer Vision, Image Processing
Computer Vision project for Intelligent Traffic Light Control implementing the concepts from Image Processing, using MATLAB.