Established Fashe, an online shopping portal featuring a wide range of fashion products. Developed a recommendation system for selected products and implemented automated invoice delivery to users' emails upon purchase.

Developed a rapid COVID-19 detection website by training a Convolutional Neural Network with sample images, achieving high accuracy. Additionally, managed a real-time visualization of global COVID-19 statistics, displaying active cases, deaths, and recoveries on a world map.

Built an application by programmatically extracting and analyzing five years of historical Bitcoin data leveraging the TwelveData API, and preprocessed it for quality and consistency. Trained LSTM neural network model to make informed decisions achieving a remarkable 78% accuracy rate in the crypto market.

Developed a predictive model to forecast loan approval, achieving an accuracy of 88%. Identified crucial KPIs such as credit score trends and debt-to-income ratios to optimize lending decisions. Designed a Power BI dashboard to visualize insights on customer segmentation enabling data-driven recommendations for operational efficiency.

Developed an advanced spatial query algorithm that outperformed traditional methods in high-dimensional data analysis. Optimized for various applications, including healthcare and urban planning, with comparative evaluations using E2LSH.

This project integrates advanced record linkage techniques to cluster healthcare records and efficiently analyze cost savings by identifying overlapping medical procedures (Blood tests, X-rays, Dental treatments) within specified time thresholds. It automates the cost-saving analysis and outputs the results to a file, offering a streamlined solution for healthcare cost optimization.

Tracking your F1-OPT status just got easier! This Chrome extension tool will help F-1 OPT students manage their 90 days of unemployment for compliance. It provide a clear overview of your remaining days by providing a visual countdown to 90-day OPT end date. The extension also offers a dark mode option for user preference.

Developed a secure, full-stack password manager using React, Tailwind, Express, and MongoDB, enabling users to store and manage encrypted credentials with authentication and intuitive UI.

Built a responsive Netflix UI clone using HTML, Tailwind CSS, and JavaScript replicating the original layout and features like dynamic content cards, banner, and navigation for an authentic user experience.

Created a Spotify-inspired music recommendation system using NLP techniques like TF-IDF and content-based filtering. Preprocessed song metadata, extracted semantic features from descriptions, and computed similarity scores to recommend personalized tracks based on user input.

Developed a distinguished mini-game at the HaXploit 24-Hour Hackathon, organized by the CS and IT Department. This event drew participants from various colleges, and the project received recognition from board members for its innovation and execution.

Published a paper after a peer review at the CiiT International Journal of Digital Image Processing sponsored 7th National Conference on Advancement in Information Technology (NCAIT), Bengaluru. (March ’21)

Successfully presented an innovative augmented reality (AR) virtual clothing try-on system to a group of potential investors. The presentation showcased the system's cutting-edge technology and user-friendly interface, securing strong endorsements and enthusiastic support from the investors.