Machine learning is infused in all walks of life — and in a lot of Google products, including Google Home, Search, Gmail, and more, and in systems such as those used by self-driving cars and fraud detection. A tremendous amount of effort is being made to improve people’s experiences using products throughout the industry, where products are powered by ML and AI. However, developing and deploying high-quality, robust ML systems at Google’s scale is hard. This can be due to many factors, including distributed ownership, training serving skew, maintaining privacy and proper data access controls, model freshness, and compatibility.
In her @Scale talk, Manasi Joshi, Director of Software Engineering at Google, discusses how Google started an ML productivity effort to empower developers to move quickly and launch with confidence. This effort encompasses building infrastructure for reliability and reusability of software, as well as the extraction of critical ML metrics that can be monitored to make informed decisions throughout the ML life cycle.