Educator and Senior Data Scientist. I teach programming, algorithms, databases, ML, and analytics, and I build production-grade models and data products that people actually use.

I currently teach at University of British Columbia (Adjunct, Land & Food Systems), with prior terms at Douglas College, BCIT, VIU, and North South University. My classes are project-based with reproducible labs, versioned datasets, and clear rubrics - so students leave with working code, not just notes.

My research sits at the intersection of anomaly detection, representation learning, and probabilistic/Bayesian modeling. I’ve published on disentangled VAEs and unsupervised detection (IEEE BigData; Machine Learning with Applications) and contributed to health analytics as an NSERC CREATE VADA trainee. As an NSERC CREATE VADA trainee I built a hypertension risk model using large population health data and delivered reproducible results with strong governance. Previously, I worked with large, messy datasets and the privacy and audit requirements that come with them. Interests include cancer imaging and brain tumor use cases where classification and segmentation can move the needle for planning and reporting.

Industry side: I’ve shipped forecasting and analytics systems in Medical labs, Clinics, pharma and real estate. Highlights include standardizing KPI layers, cutting forecast error by ~12%, reducing incidents ~50%, and improving p95 latency ~35%. 

Day-to-day tools: Python, PyTorch, scikit-learn, SQL, FastAPI, Airflow/PySpark, Docker, GitHub Actions, and Power BI/Tableau.

I enjoy turning messy requirements into stable pipelines, well-tested APIs, and readable docs. If you’re working on ML for health, public sector analytics, or data-rich education projects, let’s talk.