Writing
Series
ML in Research & Production
Machine Learning (ML) in Research and Production
Opens the series by unpacking why ML models behave differently once they leave the research environment — examining assumptions that hold in academic settings but break down under real-world production constraints.
Read on Medium →Machine Learning (ML) in Research and Production: Data Requirements
Continues the series by digging into how data requirements shift between research and production ML — covering volume, quality, and latency considerations researchers rarely confront until deployment.
Read on Medium →Understanding Data in Production Machine Learning Systems: A Visual Guide for Practitioners
Closes the series with a visual breakdown of how data drift, validation, and monitoring shape the day-to-day reliability of production ML systems — aimed at practitioners building and maintaining live pipelines.
Read on Medium →Essays
Standalone Reflections
Feb 23, 2024
Impact-Driven Research — Does Research Have to Convey Practical Contributions?
A reflection on the growing expectation for academic research to demonstrate tangible real-world impact, and what that pressure means for how research questions get framed and pursued.
Read on Medium →