Projects
Share3D
A platform for requesting and sharing models for 3D printing.
Implemented interactivity with 3D models using WebGL and made it responsive using tailwindCSS.
Leveraged MongoDB data indexing to ensure the response time of fetching GLTF/STL data within 1 second.
Stored GLTF/STL data in MongoDB with file encryption and optimized data indexing.
ProFlow
A project management tool for dragging and dropping tasks (Trello clone).
Implemented user authentication, card view, and profile features of the app using TypeScript.
Improved drag-and-drop UX with predictive transitions, eliminating database lag for smoother card movements.
Worked with a team of 4 to build a full-stack project management application.
Customer Churn Prediction
A machine learning project to predict why customers might leave a company.
Developed a machine learning model to predict customer churn reasons using RandomForest and KNN classifiers.
Achieved 92.9% accuracy on the training data and 87.1% accuracy on the test data.
Enabled Excel/CSV uploads with Pandas for automated parsing, cleaning, and rapid analysis.
Integrated the model into a Flask web app, delivering real-time results via a React front-end.
FilmIt
A movie recommendation system.
Developed a movie recommending service, filtering through IMDB API and displaying trailers through Youtube API.
Created a high-performance, reusable hash table to optimize data storage and additional API implementations.
UDP Client-Server
A course project to implement a UDP client-server application that allows users to send and receive messages.
Designed a UDP-based client-server application with client-side and server-side components using C++.
Integrated cryptography functionalities, including nonce generation, key exchange, and signature verification.
Developed a handshake protocol to establish a secure connection between the client and server using public-key cryptography, digital signatures, and certificate verification.
Compiler Optimization
A course project to optimize a given compiling algorithm.
Reduced compilation time by 70% by refactoring C/C++ compiler algorithm for optimized performance.
Optimized sorting algorithm, increasing performance by 50% when processing large data sets of over 1 million entries.
TimeCapsule Bot
A telegram bot that allows users to send messages to their future.
Created a Telegram bot that preserves memories and sends them in the future using asyncio and apscheduler.
Implemented a robust scheduling algorithm, handling over 500 future messages daily without failures.
Developed a secure and efficient message encryption system, ensuring that messages are only accessible to the intended recipient at the specified future date.