Serial entrepreneur who enjoys creating scalable new technology that serves underrepresented markets and helps people be more productive through automation. Currently building dexterous robots with the world's most advanced tactile sensors for tasks that previously required human operators. Architect of multi-task robot systems that combine high-dimensional sensor input modalities with state-of-the art deep learning approaches to perception and control.Formerly at UC Berkeley studying Information and Data Science, with a focus on deep learning applied to applications in computer vision, time series prediction and control tasks. Past projects have included data-driven forensics on video altered by artificial intelligence, evolution strategies for bio-mechanical control and generative networks. Areas of interest include causal models, reinforcement learning and meta-learning for deployed hardware operating in the wild. Areas of graduate-level study include probability and statistics, research design, field experiments, causal inference, time series analysis, information theory, deep learning, and machine learning at scale.