As OEMs move from intense R&D to scaled production for advanced self-driving systems, the key to continuously improving system performance is creating a data engine. Data engines leverage the scale of production fleets to detect system failures in real-time and iteratively improve ML performance.
In this talk, Ben Levin and Kate Park will discuss industry-leading data engines, how to enable data engines with modern production sensor suites, and how leading OEMs can leverage data engines to reach production quality in complex driving scenarios.
Kate Park is Scale's Director of Product Management focused on labeling and data engine platforms. Previously she was the first Product Manager for Tesla Autopilot's AI team where she built its data engine. Kate received her bachelor's degree with distinction at Stanford in computer science in the artificial intelligence track (Phi Beta Kappa, Tau Beta Pi, Stanford F. E. Terman Award). She has published award-winning research on spoken natural language processing, U.S. patents with Elon Musk, and a travel memoir.
Ben is Head of Enterprise Annotation at Scale AI, where he leads teams focused on 2D, 3D, and Mapping for leading L4/L5 autonomy programs, OEMs, and Tier 1s. Previously, he was General Manager for Navigation Data at Mapbox, where he led the creation of an ML-based mapping system that powered location for more than 500 million users. He also was previously a technology strategy leader at the Boston Consulting Group.