My Projects
A collection of data science projects showcasing machine learning, analytics, and visualization skills
Credit Card Fraud Detection
A machine learning project to detect fraudulent credit card transactions using real-world data. Preprocessed 100K+ records, addressed class imbalance with SMOTE, reduced dimensions using PCA, and trained XGBoost and Random Forest models for high precision and recall.
Role Radar: Find Domain Experts
A Flask-based web tool that identifies professionals with 10+ years of experience in a specific domain using GitHub and web search data. Combines Serper API, GitHub API, and Gemini API to extract, analyze, and rank profiles using a custom NLP-based confidence scoring algorithm. Delivers accurate results in under 20 seconds — no dummy profiles, only real experts.
UrbanFlow: AI Traffic Management
A real-time intelligent traffic control system using YOLOv11 for multi-class vehicle detection and AWS for scalable deployment. Trained with a Kaggle dataset (mAP@50 = 94.6%) to detect vehicles and pedestrians under varying conditions. The system uses AWS Lambda, DynamoDB, and S3 to dynamically manage signal control based on live road usage.
GenAI Learning Hub
An AI-powered, inclusive, and immersive educational platform aligned with UN SDG 4. Features AI-generated video lessons, Gemini-powered chatbot support, secure Firebase authentication, and a responsive, intuitive UI for underserved learners.
Online Cab Reservation System - Star Schema
A dimensional modeling project for an Online Cab Reservation System using MySQL and a Star Schema. Includes fact and dimension tables for analyzing reservations, revenue, driver performance, and customer behavior.
Inventory Monitor
A streamlined web application for inventory management, built with Next.js and hosted on Vercel. It offers a clean interface to track stock levels, manage product details, and monitor inventory changes efficiently. Designed for simplicity and usability, making it ideal for small businesses or personal use.
