Artificial Intelligence for Human Behavior.

Pebble Health is an all new built-in health tracking application available to the Pebble Time family of smartwatches. Developed by Pebble in collaboration with researchers at Stanford University, Pebble Health will track your daily activity and sleep patterns.My role on the Pebble Heath team has been to help improve our algorithms for activity and sleep tracking, analyze patterns of user behavior, designing scaleable data pipelines for efficient movement of data across multiple platforms, and to work with the product team to design novel features centered around user engagement and behavior change.

Fitnet (http://fit.net) is a convenient and easy virtual fitness studio that allows you to connect to a personal trainer from your mobile device.My roles at Fitnet were to design and implement a computer-vision algorithm for real-time workout assessment, aid in product design using the latest psychological research on behavior change, apply time-series analysis to user behaviors to improve product features, build in-house business intelligence data visualization systems, and manage collaborations with outside academic and health research groups.

My research focused on the processes of human decision making and the ability of changes in context to shift our decision-making process. Understanding the role of context is particularly important when we attempt to apply knowledge gained in a lab setting to the real world. I applied statistical modeling of human goal-directed behavior and functional magnetic resonance imaging (fMRI) to understand the neural networks involved in computing subjective value and their dynamic modulation by context.

During my PhD I designed, executed, analyzed, and wrote a thesis on auditory streaming and auditory attention. Auditory streaming is the phenomenon where a complex auditory scene with overlapping sources is broken down into individual objects. After this process is complete selective attention is involved in preferentially processing the information from a single stream. In my research I identified novel EEG correlates of auditory streaming, used fMRI to support the hypothesis that auditory streaming occurs in primary auditory cortex, and used fMRI to show the shared and distinct attentional networks involved in attending to an auditory stream based on either pitch or location cues.