Applied machine learning

Representing DataKind, I led a team of volunteers to analyze and classify text data for Empower Work, a start-up that provides free counseling services to distressed employees. Our deliverables included our insights on what contributes to a successful conversation and a web app that predicts labels given the messages to alleviate the human labeling process.

At Uber, I focused mainly on the area of NLP, in particular, dialog systems. Before then, I worked primarily in general ML projects and product analytics.Select experience:- Built chatbots to assist Uber agents in helping drivers navigate through the onboarding funnel.- Built a voice assistant for drivers to respond to rider messages vocally in order to reduce the accident rate.- Built rider targeting models to boost user engagement and increase the company's market positions.- Analyzed and segmented rider behaviors including their usages, preferences, and evolutions.- Built data pipelines and dashboards to track KPIs, experiment results, and model performance.

Developed statistical models in support of expert testimony and academic research. Typical projects included estimating the impact of a special event on a company’s financial outcome using econometric models.

Sell-side M&A

Assisted professors on research projects ranging from mobility management to network resource management.