Welcome to my work sample!

What you should take away: I understand how to program well and construct end-to-end data science projects. I like to think of this work sample as the extra credit to the coding interview.

What this is not: These projects, while I find them interesting, are not novel. Novel projects take orders of magnitude more time, money, data, and are outside the scope of what I wanted to accomplish.

If you take a look around and would like to get in contact with me, check me out on LinkedIn.

Magic: The Gathering (MTG) Artist Recognition

Using PyTorch and Azure ML I train a convolutional neural network for recognizing MTG artists. I containerize my models using Docker, and deploy it as a web application. Azure Devops pipelines are used to run data pipelines and automatically retrain models.

Azure ML Diabetes Classification

Using Docker, AzureML, XGBoost, Hyperopt, DVC, Poetry, and other related tools I create and deploy a model for classification of diabetes using BRFSS data from the CDC.

John Cummings's Work Sample

Recent Posts

Magic: The Gathering (MTG) Artist Recognition

Azure ML Diabetes Classification