Experience

Project Manager of Agentiiv's MCP Gateway ML project, building a production-grade communication platform for AI agents and MCP servers, as part of a UTMIST industry collaboration. Defined project requirements, scope, and timelines, while tracking progress and managing the team using Agile Scrum and Jira. Leading system design and implementation, as well as addressing technical challenges.

Developing AuditMate, an AI-driven web and mobile platform that automates building energy audits for BuildingAssets, as part of a UTMIST industry collaboration. The application analyzes energy audits with object detection, OCR, and LLMs.

Performing data analysis and image preprocessing. Building and integrating machine learning and deep learning pipelines that support both professional auditors and individual clients, enabling self-serve and guided audit experiences.

Deploying the full cloud-based architecture using AWS EC2 and integrating AI services with Next.js and Flutter frontends to deliver scalable, accessible applications.

Full-stack developer and organizer of GenAI Genesis 2026, Canada’s largest AI hackathon with 2000+ applicants, 800+ hackers (30% increase from 2025), 250+ projects, and 90+ judges. Built and maintained the website, as well as participant and judging platforms, which supported 10,000+ uses during the hackathon.

Designed and implemented secure REST APIs with Zod validation, database schemas, role-based access control, and user interfaces, as well as CI/CD pipelines with GitHub Actions and Vercel for scalable deployment. Collaborated with cross-functional teams, delivering a seamless hackathon experience.

Leading development of the SceneClarity ML project, an interpretable deep learning pipeline for autonomous vehicle perception, and managing the project using Jira. Analyzing and augmenting large datasets, and building ML models for object detection (YOLOv11, LaneNet), classification (MobileNetV3, ResNet-50), and unsupervised cause attribution (KMeans, GMM).

Integrating outputs from multiple models and researching scoring and weighting methods to reflect real-world autonomous vehicle safety. Delivering Dockerized REST APIs and a React web app for use by other researchers and developers. Presented the project to stakeholders at various events and conferences, including the CUCAI 2026 Conference and the EigenAI 2025 Conference.

Built integrated bioinformatics and machine learning pipelines for the Functional Gene Expression project, enabling simultaneous benchmarking of 15+ models and achieving 3× faster analysis. Analyzing multi-dataset gene expression data to identify disease biomarkers and predict autoimmune diseases, training linear, ensemble, gradient boosting, and SVM classifiers, with SHAP/LIME for interpretability. Implemented Airflow + dbt ETL pipelines to transform and store GEOparse data in a Supabase PostgreSQL database.

Official contributor and modder of the starblast.io game. Developed the official “Capture the Flag” mod, which has been played over 2 million times, improving performance, balance, and engagement. Developed multiple mods using the Starblast.io API, implementing real-time game logic in JavaScript with WebSockets, and designing custom 3D ships using CoffeeScript for Three.js rendering.

Mentored and led FTC teams 16488 and 22101. Designed and led the development of PID-based motion control and finite state machine architectures for autonomous robot operation. Implemented real-time object detection using TensorFlow models integrated with OpenCV for vision-based decision making. Developed high-fidelity robotic simulations directly interfaced with the FTC SDK to validate control algorithms and model complex robot behaviors. Won 2nd place at the Ontario Provincial Championships, as well as Innovate and Design Awards.

Designed and developed the web platform using the SvelteKit framework. Engineered intuitive database schemas and relationships in the Firestore cloud database to support core app functionality such as user management and course enrollment. Continuously improved the user experience based on user feedback.

Led, trained, and supervised a team of tutors to teach students in programming, STEM, and languages. Adjusted course trajectory based on student feedback. Developed marketing initiatives to expand the organization’s reach, positively impacting 300+ students across 50+ schools in Ontario.

Part-time tutor at Upper Markham Learning Centre, teaching coding, robotics, competitive programming, and high school STEM subjects including math and physics. Delivered personalized lessons, managed multi-level classrooms, and provided detailed progress feedback to parents while adapting strategies to support individual learning needs.

Led the coding section at the non-profit organization FutureEd 4 Kids, managing mentors and organizing team meetings. Designed new programming courses to expand the coding section in the organization. Taught students in engaging and interactive classes.