Created a variety of highly interactive custom graphics for the IsoPlexis software package for use by external customers. Designed and developed a complex statistical suite that easily guides users through utilizing the appropriate statistics. Built an image classification system for detecting RNA barcodes for mapping single-cell proteomics and mRNA data. Improved the predictive accuracy of the single-cell secretome pipeline with novel machine-learning and deep-learning techniques.

Created a model for predicting fashion trends using a combination of deep-learning, clustering, and time-series analysis. Designed a tool for optimizing labor allocation in stores based on sales, customer traffic, and various employee tasks. Developed data wrangling and munging tools to increase analytic efficiency across subsidiaries and e-commerce and brick and mortar.

Data-mined millions of scientific studies that pertain to the products Cognitive Genomics offers. Utilized cutting-edge text-mining tools to search for genetic mutations that relate to cognitive function and extracting the relevant information to produce customized reports. Created complex models for predicting cognitive phenotypes, and developing comprehensive tailored advice for customers based on their genotypes.

Commercial insurance predictive analytics, specializing in text-mining.Created a predictive model for detecting recovery (subrogation) opportunities for Worker’s Compensation claims. Used natural-language processing to analyze claim diaries, which supplemented the use of various structured data. Developed spell-correction and abbreviation-disambiguation tools and other in-house algorithms for cleaning and structuring textual data. Designed a model that predicted which motor-vehicle-accident claims involving head injuries were likely to result in an allegation of Traumatic Brain Injury (TBI). In the process, helped develop text-mining algorithms for broader use within the company. Later, the TBI model was enhanced to include a self-generating report that notifies specialists to investigate, collects feedback on actions taken, and helps monitor the financial impact of the intervention.

Devised novel experimental paradigms and data-analysis methods, leading to the first evidence demonstrating how and where the human brain accumulates information to facilitate ambiguous decisions. (See publication information below.) Developed computational models of decision-making behavior and its neural correlates. Mentored PhD students on projects that demonstrated how confidence evolves during a decision, using psychophysical tools and Bayesian decision theory.

Researched genetic components of financial decision-making, focusing on risk aversion.

Oversaw undergraduate research leading to a better understanding of perceptual drivers of voterpreference