Resume

Principal Data Scientist

Download Resume (PDF)

Professional Summary

Principal Data Scientist with 10+ years of experience designing, building, validating, and deploying production-grade machine learning, optimization, forecasting, and AI solutions across retail, healthcare, and public-sector environments. Strong hands-on experience owning end-to-end analytical workflows — from data ingestion, exploration, feature engineering, and model development to output validation, deployment, and stakeholder-facing insight generation.

Technical Skills

Programming & Analysis
Python, SQL, PySpark, R, Java, SAS
Machine Learning / AI
Scikit-learn, XGBoost, CatBoost, LightGBM, Keras, AutoML, Recommender Systems, Reinforcement Learning
Data & Cloud Platforms
Databricks, Spark, Azure, Ray, MongoDB Vector Search
MLOps / Delivery
MLflow, Git, GitHub, Azure DevOps, Streamlit
Optimization / Decision Science
Gurobi, CPLEX, Simul8, Simio, Arena
Generative AI / Applied NLP
RAG, Agentic RAG, LangChain, Agents SDK, Prompt Orchestration, LoRA/QLoRA, LLM-as-a-Judge

Professional Experience

Senior AI/ML Engineer (Senior Data Scientist)
Canadian Tire Corporation
Feb 2022 – Present · Ottawa, ON
  • Led the development of recommender systems for loyalty customers using POS transaction data and promotional calendar data, driving more relevant weekly offer personalization and improving conversion.
  • Owned the development and deployment of a production ML model to identify false in-stock signals, turning a high-impact retail inventory challenge into a scalable inference solution that helped prevent approximately $13M in annual lost sales.
  • Led the engineering of reusable ETL and feature pipelines in Spark and SQL, enabling scalable model training, inference, KPI reporting, and faster iteration across retail ML use cases.
  • Designed and implemented a decision-optimization system in Python and Gurobi to improve staff scheduling and operational planning, delivering approximately $1M in annual savings.
  • Built a discrete-event simulation model to optimize labor allocation across retail stores, increasing average net earnings by 5.5%.
  • Designed and prototyped an agentic RAG-based AI system using MongoDB Vector Search, MLflow, and Streamlit to improve product discovery and enable richer customer interactions.
  • Fine-tuned Mixtral, BERT, and Flan-T5 models using LoRA/QLoRA to classify cloud-migrated data tables, improving metadata classification efficiency.
  • Developed automated evaluation workflows, structured error analysis, and LLM-as-a-Judge approaches to assess output quality and support systematic iteration across prompts, retrieval, and model behavior.
  • Mentored junior data scientists across project delivery, code quality, feature engineering, model development, and stakeholder communication.
Research Assistant / Lecturer
Telfer School of Management, University of Ottawa
Sep 2015 – Jan 2021 · Ottawa, ON
  • Developed an advanced analytics solution combining discrete-event simulation, GLM, and inverse optimization to redesign clinic operations; reduced total patient wait time and overtime costs by 60% at The Ottawa Hospital.
  • Implemented a reinforcement learning–based decision system using value function approximation and column generation to optimize healthcare capacity allocation.
  • Designed a multi-criteria decision analysis framework to support formulary decisions for public and private drug plans.
Data Scientist
Bank of Canada
Sep 2019 – Aug 2020 · Ottawa, ON
  • Built production-ready time series forecasting models using STL, dynamic harmonic regression, TBATS, Random Forest, and LSTM to support capacity planning and demand forecasting, improving forecast accuracy by 10–15%.
Data Scientist (Health Economics)
Pivina Consulting Inc. / Ontario Health
Sep 2012 – Jun 2014 · Ontario
  • Developed health economic evaluation models using decision trees, Markov chains, cohort simulations, and SAS to support evidence-based reimbursement decisions.
Research Assistant
Department of Epidemiology & Biostatistics, Western University
Sep 2010 – May 2012 · London, ON
  • Built a cost-effectiveness model using survival analysis, Markov models, cohort simulation, SAS, and Ontario health datasets, informing a reimbursement recommendation in the Canadian healthcare system.
  • Developed a discrete-event simulation model to optimize staffing and reduce wait times for radiation treatment, improving patient throughput and resource utilization.

Certifications

Education