A long-standing grand challenge in computing is to enable machines to act autonomously and intelligently: to rapidly and repeatedly take appropriate actions based on information in the world around them. Driven by trends in the data economy, rapid progress in AI, and an increasingly programmable physical world we are at an inflection point that demands a new class of AI system. This new class of systems goes beyond training models at scale, to connecting models with the world, rendering predictions in real-time under heavy query load, adapting to new observations and contexts. These systems will need to be composable and elastically scalable to accommodate new technologies and variations in workloads. Operating in the physical world, observing intimate details of our lives, and making critical decisions, these systems must also be secure.

In this talk, I will present work in my group exploring advances in systems for prediction serving, autonomous driving, and how model design and system design interact.  In particular, I will discuss some of the key trade-offs between time, accuracy,  convergence, throughput, and security that govern how we design systems, train models, and make predictions that meet the demands of real-world applications.