A browser-based pixel art editor built with React and Vite that lets users design, edit, and explore artwork at the pixel level using real-time tools on a responsive canvas. It supports a dynamic grid canvas (resizable, scrollable for high resolutions), common tools like pen, eraser, fill, color picker, and shows status info such as the active tool, color, canvas size and dimensions. Key ambitions also include features like undo/redo, image import/export (PNG), possible animation support, and mobile friendliness.

An autonomous microagent that automatically summarizes YouTube lecture videos, processing 1+ hours of content to generate concise, accurate summaries along with quizzes to test knowledge. The system leverages speech-to-text transcription and natural language processing to improve study efficiency and knowledge retention for the users.

An autonomous microagent that parses source code and generates line-by-line natural language explanations using Llama3. Capable of handling files in Python, it streamlines codebase understanding for students and developers. Built using Streamlit and Python, it processes 1000+ lines of code per session and is optimized for fast, local interaction.

A desktop-based smart CCTV system with a Tkinter GUI, capable of detecting motion in real-time using OpenCV and capturing intruder snapshots. Designed for lightweight local surveillance, it processes 30+ FPS video streams, auto-saves suspicious frames, and allows manual or automated camera control. Built entirely with Python, this tool is ideal for low-resource environments needing intelligent monitoring.

A machine learning-powered web application for predicting diabetes risk based on medical input features such as glucose level, BMI, age, and insulin. Trained on the PIMA Indian Diabetes dataset, the system achieves 85%+ accuracy using models like Logistic Regression, Random Forest, and SVM. Built using Python, scikit-learn, and deployed via a Flask web interface with a clean UI for real-time risk assessment.

A full-stack task management system featuring Role-Based Access Control (RBAC) for Admins, Managers, and Employees. Built using React.js, Node.js + Express, and MongoDB, it supports real-time task assignment, status tracking, and secure authentication. It is designed for small team productivity, the system handles 100+ concurrent tasks and includes features like routes, JWT auth, and modular APIs.

A generative art engine that programmatically and automatically creates 1000+ unique NFT assets by layering traits across multiple categories such as backgrounds, accessories, and characters. Built with Python and PIL, it ensures trait rarity control, uniqueness validation, and automatic metadata generation for blockchain deployment. Ideal for bulk minting on platforms like Ethereum and Solana.