• Building SOTA neural networks leveraging the YOLO framework for real-time object detection and instance segmentation; e.g., built a drone-based detection model that achieved world-record performance (mAP50-95) on the VisDrone dataset.• Building two PERN web apps (Postgres, Express, React, Node), integrating ML pipelines and optimizing KPIs.

• SYSC 2510: Probability, Statistics, and Stochastic Processes (Python, R)• SYSC 2006: Imperative Programming (C)• ECOR 1042: Data Management (Python)• SYSC 3101: Programming Languages (Prolog, Racket)• SYSC 2004: Object-Oriented Development (Java)

Built the Data Protection on Demand (DPoD) platform to provide a wide range of cloud hardware security module and key management services through an online marketplace, making security simpler, more cost-effective, and easier to manage. DPoD's microservice architecture is based on a variety of technologies, including Java/Spring, Python, Go, TypeScript/Jasmine, Protractor, REST and Swagger, Concourse, Pivotal Cloud Foundry, AWS, and Angular for the frontend. The SAFe framework was utilized to align multiple teams on common objectives, facilitating our continuous enhancement, delivery, and monitoring of the multi-site application with zero-downtime deployment.

Built the Machine Trust Platform (React-Redux frontend, Python/Flask-SQLAlchemy and Java/Spring Boot microservices, Postgres RDS, integrating Kafka and ELK stack) for measuring, monitoring, and building trust in ML algorithms, integrating various AI trust and governance tools, e.g., IBM's AI Fairness 360 and the Government of Canada's Algorithmic Impact Assessment. Also built a C#/ASP.NET SaaS for providing multidimensional measures of the trustworthiness of information and technologies managed and overseen by governments in emergency situations such as the COVID-19 pandemic.

• Doubled biodiesel feedstock production to 15 tonnes/day within a year using statistical process control, boosting revenue by $1.4M/year; optimized automation & data acquisition systems (e.g., distributed control system, programmable logic controllers) and control schemes. • Tuned PID controllers using MATLAB/Simulink, and optimized control strategies using the ArchestrA InFusion Integrated Engineering Environment. • Troubleshoot lead for the biodiesel feedstock recovery process. Engineering resource in root cause analyses. Revised standard operating procedures.