01.Traffic Fatality ML Model

Developed a machine learning model using Toronto Police’s Kill and Seriously Injured dataset to predict traffic collision fatalities with 80% recall, featuring a Flask-based API for real-time predictions.

Python, Flask, Toronto KSI Dataset

02.Live Aquaria

AI-powered app that identifies fish species using Google’s Gemini API. Includes logging features, an interactive 3D aquarium with React and Three.js, and a global fishing map via Leaflet API.

React JS, Node.js, Express, MongoDB, Three.js, Leaflet API

03.Email Spam Detection - NLP

A Natural Language Processing project to classify text comments into spam or non-spam categories using Python. Implements preprocessing, feature extraction (TF-IDF), and Naive Bayes-based classification with hyperparameter tuning.

Python, Scikit-learn, NLP, TF-IDF

04.FX RouteDesk

Built a multi-leg FX routing tool that ranks the top three fiat and stablecoin routes by final recipient amount, with per-leg fees, direct-route comparison, scale analysis, and provider health handling.

Next.js, TypeScript, React, Vitest, Recharts, React Flow