Unverified60% confidenceFactExact time
Google's machine learning engineers have summarized from experience that actual machine learning code in a machine learning system typically accounts for only 5%, while the remaining 95% consists of data pipelines, service wrappers, and monitoring systems.
2
Sources
60%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
An Efficient Learning Guide for Data Science: Study Partnerships and Hands-On ML in Practice
redditr/learnmachinelearning7/11/2026
Related Claims
VerifiedGoogle's 2015 paper "Hidden Technical Debt in Machine Learning Systems" revealed that the actual machine learning code in ML systems accounts for only a very small proportion.79% similarUnverifiedGoogle Professional Machine Learning Engineer 认证难度公认较高，覆盖范围广，更强调机器学习的系统设计能力77% similarVerifiedIn the paper "Hidden Technical Debt in Machine Learning Systems," Google pointed out that in real ML systems, the actual ML code constitutes only a small fraction of the entire system, surrounded by large amounts of data collection, feature extraction, configuration management, serving infrastructure, and monitoring code.77% similarUnverifiedGoogle claims to use machine learning for ad content moderation.71% similarUnverifiedGoogle lists problem definition as the foremost principle in its Rules of Machine Learning (43 rules).70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/491637API
curl https://kongchang.com/api/v1/knowledge/claims/491637MCP
get_claim(id=491637)