Explore the critical role of ML Observability in the machine learning stack through this 44-minute conference talk from MLOps World: Machine Learning in Production. Delve into common model failure modes, including model drift, data quality issues, and performance degradation. Discover how ML Observability addresses these challenges by implementing monitoring systems, providing troubleshooting tools, and identifying root causes. Learn about best practices and industry examples that showcase the importance of ML Observability in creating a feedback loop for continuous model improvement. Gain insights from Reah Miyara, Head of Product at Arize AI, who brings extensive experience from Google AI, IBM Watson, Intuit, and NASA Jet Propulsion Laboratory to this discussion on enhancing machine learning systems' reliability and performance.
ML Observability: A Critical Piece in the ML Stack
MLOps World: Machine Learning in Production via YouTube
Overview
Syllabus
ML Observability: A Critical Piece in the ML Stack
Taught by
MLOps World: Machine Learning in Production
