Data Science Consulting

Quome

January – June 2025

At Quome, I mentored a team of data scientists developing AI agents using Quome’s secure, cloud-native infrastructure. Quome enables organizations to develop and deploy AI applications within their own cloud environments, with an emphasis on data security, isolation, and regulatory requirements in industries such as healthcare. My work focused on building scalable, production-ready AI solutions while guiding the team’s technical development and engineering practices.

I also led a comprehensive refactoring and restructuring of the codebase to improve maintainability, readability, and long-term scalability while reducing technical debt. Additionally, I developed technical training materials and established engineering standards to promote consistent coding practices, reproducibility, and high-quality AI software development across the team.

Veda Data Solutions (now H1)

April – December 2023

Through the Erdős Institute, I led a team of data scientists in conducting an independent, third-party validation of Veda’s AI-powered provider data curation platform. We evaluated the accuracy and representativeness of their provider directory data compared with traditional healthcare provider directories, assessing its potential to improve data quality, reduce administrative burden, and help patients access care.

I also developed a statistical evaluation framework to assess healthcare provider network coverage, identify gaps in representation, and recommend improvements. Our findings contributed to a published white paper and press release, highlighting the importance of independent AI validation in healthcare and the potential for automated data curation to improve provider directory accuracy and transparency.

Manganese Health Data Solutions

March – May 2025

This project was a collaboration with VillageMD, a healthcare network that provides value-based care for patients. In this role, I worked on simulating scenarios to understand the dynamic relationship between (1) patients with different care needs, (2) healthcare operations, such as clinician allocation, and (3) the economic portfolio of the healthcare provider. The goal was to use data science techniques to optimize care for each patient and ensure that the value-based care practice runs smoothly and efficiently.