David Sculley, Gary Holt, Daniel Golovin, Eugene Davydov, Todd Phillips, Dietmar Ebner, Vinay Chaudhary, Michael Young, Jean-François Crespo, & Dan Dennison (2015), References, Textbook of AI

David Sculley, Gary Holt, Daniel Golovin, Eugene Davydov, Todd Phillips, Dietmar Ebner, Vinay Chaudhary, Michael Young, Jean-François Crespo, & Dan Dennison (2015)

Proceedings of NeurIPS 2015.

URL: https://papers.nips.cc/paper/2015/hash/86df7dcfd896fcaf2674f757a2463eba-Abstract.html

Abstract. An influential paper from Google arguing that ML systems incur significant and often hidden technical debt in data dependencies, configuration, feature pipelines, monitoring, and infrastructure, with the model itself typically a small fraction of the total system.

Tags: mlops technical-debt production url-only

AI tools used: Claude (research, coding, text), ChatGPT (diagrams, images), Grammarly (editing).

Hidden Technical Debt in Machine Learning Systems