My Projects
A curated selection of my professional endeavors and personal explorations in software and data.
Professional Work
Triumph College Consulting Website
Built a full-stack website featuring dynamic content management and responsive design. The site includes features for student success tracking, scholarship management, and consultation scheduling.
Key Achievements
Enhanced user experience through intuitive UI/UX design
Implemented streamlined contact interfaces
Developed dynamic content management system
AI-Powered Assignment Grading System
Developed a comprehensive grading system integrating AI capabilities with a robust full-stack architecture. The system features secure data handling and efficient processing of student assignments.
Key Achievements
Improved grading speeds by 75%
Increased user satisfaction by 20%
Enhanced system status visibility by 15%
Improved error messaging clarity by 25%
Personal Projects
CanTrip – Canadian Travel Planning Platform
Building a full-stack travel planning platform that uses LangGraph AI agents to generate personalized itineraries, packing lists, and travel recommendations for destinations across Canada.
CoverForMe - AI Cover Letter Generator
Engineered a full-stack application using React and FastAPI that automatically generates tailored cover letters, reducing writing time by 80%.
Shyft RAG App – Retrieval-Augmented Generation System
Built a full-stack RAG application with React and FastAPI, enabling real-time question answering over user-uploaded documents with sub-second response times.
Heart Disease Patient Clustering Analysis
Applied K-Means clustering to segment 303 heart disease patients into 3 distinct profiles, achieving a silhouette score of 0.1908 and uncovering patterns for personalized treatment.
Smartphone Price Prediction Analysis
Designed and implemented end-to-end ML pipeline achieving 96.5% prediction accuracy. Engineered robust data processing system reducing feature extraction time by 40%.
Automated Fake News Detection
Engineered data pipelines and implemented LSTM, RNN, and BERT models for fake news detection, achieving high accuracy on various datasets.
ForumRank Website
Collaborated on a custom-designed discussion post website with contextual menus, responsive design, and MySQL database integration.
