This repo holds the code and slides for a talk that I gave at the Alteryx Inspire 2018 conference about deploying a Python model to Alteryx Promote. The model is a NHL expected goal probabilty model.
In this repo, you'll find the Jupyter notebook that I used to train the model along with the code to deploy the model to Promote from Python.
Also in here is an example application built with Node.js and React
To train use the Jupyter notebook, you'll need the dataset (~600MB). There's a script in the /data folder to get the file and unzip it.
cd data./get_data.sh
Enjoy!