Weekly seminar: Scott Toborg
February 22, 2024
Where: TMCB 1170
When: February 29th @11am
Talk title: Ultra-Scale Intelligent Systems: A New BYU Research Area Investigating Intelligence at Scale
After seeing so much progress in AI over the last few years, it’s hard to understand why many enterprises still struggle to bring AI models into production. This research investigates the fundamental drivers of AI system development to identify roadblocks and potential solutions. To address these issues, the relatively new field of Machine Learning Operations, or MLOps, seeks to improve AI development across the entire lifecycle. These improvements promise a shorter time to market, lower costs, and higher value. Leveraging enhanced AI development processes, we explore the possibilities of scaling AI systems to support millions of models in production. At this scale, typical model metrics are no longer sufficient, and new analysis methods are needed. We present a framework for characterizing Ultra-Scale systems and discuss potential approaches for research and analysis. These Ultra-Scale systems could enable novel applications across diverse areas such as manufacturing, demand forecasting, social networks, IoT systems, traffic management, and swarm intelligence.
Dr. Scott T. Toborg is a technology leader with a proven track record of driving innovation and delivering impactful results across diverse industries. Most recently, he served as Vice Presidentof Data Science at NuSkin, where he led a digital transformation in the integrated health, wellness, and beauty industry by using AI and machine learning to improve customer experiences and outcomes and building a modern data platform for real-time IoT, AI/ML, and eCommerce. Prior to NuSkin, Dr. Toborg headed up the development of all data science and analytics products for Teradata, a leader in Enterprise Data Warehousing. These products continue to impact company revenue and helped change the company's overall focus and market position. While at Teradata, he also worked as a Principal Data Science consultant for major F500 corporations such as AT&T, Verizon, Nike, Boeing, and T-Mobile (where he co-invented a patented ML approach for wireless throughput prediction). Before Teradata, Dr. Toborg was a Principal Data Scientist in AT&T’s advanced marketing analytics group, where he developed AI-driven algorithms and systems for better customer retention (as an aside: While working at AT&T Labs in 1995, he coded the first customer churn prediction application). Dr. Toborg also served as Vice President of Products for Control4, a leading Home Automation company, and Vice President and Chief Information Officer for 360networks, the second-largest global fiber-optic services company. He has worked at four major industrial laboratories, including IBM Research, ATT Labs, MCC Research Consortium, and HRL Laboratories, along with 12 start-ups (including co-founding 2 venture capital companies). He holds a BS degree in Applied Physics (with honors), an MS degree in Computer Science from BYU, and a Ph.D. in Electrical and Computer Engineering from the University of Southern California, where his research focused on machine learning algorithms for computer vision mapped to novel massively parallel 3-D wafer-scale integration architectures. He is currently serving as an Adjunct Professor of Computer Science, teaching the Data Science Capstone class CS 483. Scott served a mission to his ancestral homeland, Hamburg, Germany, has been happily married for over 35 years, and has 3 children and 5 grandchildren. Dr. Toborg loved his formative experiences at BYU and, after a fulfilling career in industry, is excited to share his knowledge and insights to help prepare the rising generation for their turn to change the world for good.