ethanswan.com

Deploying a Model Prediction Server (Tutorial)

Conference Talk

PyCon 2023

2023-04-20

This is a 3-hour tutorial during which we start with a trained scikit-learn model and incrementally build a working FastAPI application to deliver its predictions in realtime. It’s targeted at data scientists, and no prior experience with API development is expected.

ML Engineering: From Models to Value

Conference Talk

University of Cincinnati Data Science Symposium 2022

2022-11-08

This was one of the featured presentations at UC’s Data Science Symposium. I discussed what machine learning engineering is, why companies should consciously allocate roles for it, and how to organize a combined team of data scientists and ML engineers. Unfortunately the talk wasn’t recorded, but you can view the slides using the link below.

Foundational Infrastructure to Create a Successful Data Science Team

Conference Talk

PyData Global 2021

2021-10-30

My talk with teammates Brad Boehmke and Gus Powers about what it takes to build internal tooling that effectively enables a large department of data scientists. Fairly high-level and doesn’t require too much technical background.

Python Bytes, Episode 252

Podcast Appearance

Python Bytes Podcast

2021-09-29

I was a guest on the Python Bytes Podcast to join in the week’s roundup of news in the Python world. We talked about JupyterLab Desktop, requests-cache, and a new PEP in the works to simplify the typing of decorators.

Scaling Data Science Across Python and R

Podcast Appearance

Talk Python Podcast

2019-10-29

My boss, Brad Boehmke, and I went on the Talk Python Podcast to discuss the benefits and challenges of supporting both R and Python across a large data science department at 84.51˚.