I build machine learning projects across security, healthcare, NLP, education, computer vision, and business analytics. This portfolio highlights clean project summaries, live demos where available, and direct links to the source code.
Use the filters to scan the portfolio quickly without the older carousel issues and placeholder links.
Machine learning system for malicious Android banking app detection using 50 APK-derived features such as permissions, API calls, and certificates.
RAG chatbot for medical research papers that retrieves and cites exact source passages instead of depending only on base model knowledge.
Multi-model pipeline that turns uploaded images into captions and then expands them into short stories using fine-tuned generative models.
Customer churn prediction workflow for identifying attrition risk and helping prioritize retention decisions with explainable business-facing outputs.
Sequence modeling project trained on Shakespeare's Hamlet to predict the next word and demonstrate language modeling on literary text.
Web-based prediction system for estimating student mathematics performance using academic and behavioral inputs, deployed on AWS Elastic Beanstalk.
Vehicle price estimation project built around supervised learning to predict car prices from structured features.
Prediction system for heart disease risk assessment using patient health data and a browser-based interface for quick clinical-style evaluation.
Environmental prediction project for estimating forest fire risk from weather and field-related inputs.
A concise overview so the page stays easy to scan without feeling crowded.
I focus on practical machine learning work: training models, building usable interfaces, deploying demos, and presenting results clearly.
This version of the portfolio replaces the fragile carousel layout with a cleaner card grid, fixes missing project links, and organizes the content around stronger scanning and mobile compatibility.