Professional Summary
Data scientist and cloud engineer with 5+ years of experience modernizing government digital infrastructure, blending expertise in Python/R/Julia with production-grade Kubernetes, Kubeflow, and Azure deployments. I specialize in operationalizing AI/analytics into secure, scalable systems, balancing innovation with compliance. When I joined Statistics Canada, they were still using SAS but looking to migrate to R/Python. I was instrumental during this initiative, working both in the backend on Linux, R, and Python and Kubernetes and Kubeflow engineering and also as team lead, project planning and reporting to upper management.
My background includes a Master's in Statistics with research focus on generative adversarial networks, game theory, information theory, and optimal transport. This theoretical foundation, combined with practical experience in containerizing legacy systems, designing multi-tenant platforms, and bridging technical teams, equips me to solve complex challenges in modern data infrastructure. I learned Golang during the development of the namespace-cleaner project at Statistics Canada.
Professional Experience
Designed and deployed scalable machine learning infrastructure on Kubernetes, leveraging Kubeflow and JupyterLab to optimize MLOps workflows. Spearheaded integration of SAS with JupyterLab, enabling secure migration to Python/R while improving user adoption. Built foundational Kubernetes expertise here, focusing on secure, production-grade system design. Secured operational authorization (ATO) through rigorous security protocols.
Kubernetes
Kubeflow
JupyterLab
MLOps
SAS Integration
Security Protocols
Golang
Led the design, development, and compliance of an AI-powered web application to automate health inspection report classification. Ensured adherence to F.A.I.R. data principles and federal standards while enhancing system security, data integrity, and cross-functional collaboration.
AI/ML Development
Front-End Development
Data Compliance
Security
Cross-functional Leadership
Automated Consumer Price Index (CPI) calculations via a self-developed AI clothing classification model, reducing manual effort and errors. Implemented secure, compliant data pipelines to align with federal protection standards, ensuring audit-ready processes.
Machine Learning
Data Pipelines
Statistical Analysis
Compliance
Projects
Comprehensive survey of mathematical foundations including game theory, information theory, and optimal transport. Research focused on the theoretical underpinnings of generative models and their practical applications in machine learning.
github.com/bryanpaget/generative-models
Interactive dashboard analyzing Toronto's bikeshare system with data visualization and usage statistics. Built with R and Shiny to provide insights into biking patterns and station utilization.
github.com/bryanpaget/toronto-bikeshare
Golang utility for cleaning up Kubernetes namespaces in multi-tenant environments. Developed to automate resource management and improve cluster efficiency.
github.com/StatCan/namespace-cleaner
Publications
Paget, B. (2016). "Automatic Classification of Poetry by Meter and Rhyme." FLAIRS-29, May 16, 2016.
Paget, B. (2019). "An Introduction to Generative Adversarial Networks." Master's Thesis, University of Ottawa.
Certifications
AI-900 Exam Preparation: Microsoft Azure AI Fundamentals - Microsoft Azure Machine Learning and Artificial Intelligence (AI) - Issued Apr 2023
AZ-900 Exam Preparation: Microsoft Azure Fundamentals - Cloud Computing and Microsoft Azure - Issued Apr 2023
Awards & Honors
Swartzen Memorial Scholarship - For academic excellence in the Department of Mathematics and Statistics at the University of Ottawa (Dec 2018)
Undergrad Research Opportunity Program (UROP) - University of Ottawa (Jan 2014)