Project Experiences
InCollege | COBOL, Node.js, Docker, Render, Vercel, Git, Agile (Scrum), Jira
Aug 2025 – November 2025
Developed and deployed a full-stack LinkedIn-style career networking application using a COBOL backend compiled and executed in a custom Docker environment. Integrated the COBOL system with a modern Node.js API to enable real-time account creation, login authentication, messaging, job posting, profile management, and network connections.
Built a cloud-ready architecture using Render (Docker Web Service) for the COBOL backend and Vercel for the frontend, enabling seamless cross-service communication. Implemented CORS, API routing, and automated builds that compile COBOL into a Linux binary during deployment.
Collaborated in a 6-member Agile team; contributed to core feature design, debugging across COBOL and Node, text-file based storage handling, Git branching strategy, pull requests, and sprint management using Jira.
Gaming Trends Dashboard | Tableau, Data Analysis
Jan 2025 – May 2025
Created a dynamic Tableau dashboard analyzing player engagement, revenue trends, and stream viewership.
Leveraged a Kaggle dataset to deliver actionable insights and improve report clarity by 56%.
RISC-V Pipeline Optimization | Assembly, WebRISC-V, Computer Architecture
Jan 2025 – May 2025
Implemented and optimized 5-stage pipelined RISC-V CPU datapaths to analyze performance bottlenecks and hazards.
Applied forwarding and instruction scheduling techniques to reduce stalls by ~20% and improve overall cycle efficiency by 25%, enhancing instruction throughput and execution speed.
Sorting Algorithm Visualizer
An interactive web application to help users understand and visualize how different sorting algorithms work using HTML5, CSS, and JavaScript.
Users can adjust number of bars, generate new sets, and choose from Bubble Sort, Selection Sort, Quick Sort, and Count Sort, with real-time animations and complexity info.
MNIST Handwritten Digit Identifier
Built and trained a neural network to classify handwritten digits from the MNIST dataset using Keras, with multiple dense layers and dropout regularization.
