Human Stories in AI: Amy Finnegan

Overview

Amy Finnegan, Deputy Director of Data Science at Intra Health International, details her journey from a PhD in Public Policy to a career in global health data science. Her work involves leveraging secondary data, developing R and geospatial analysis skills, and applying these to public health challenges like improving voluntary medical male circumcision program efficiency in Tanzania using Facebook population density maps and Demographic and Health Surveys. Finnegan also discusses the importance of trust-building, clear communication, and leveraging open-source tools like R and QGIS in resource-constrained environments, while also exploring the potential of LLMs like GitHub Copilot for software development.

Key takeaways

Amy Finnegan's PhD work in Public Policy at Duke involved data science methodologies (merging datasets, A/B testing) before the field was formally recognized.
A key career catalyst was recognizing the value of programming skills, particularly in R, for social scientists.
Finnegan successfully used high-resolution Facebook population density maps combined with DHS and PEPFAR data to optimize public health interventions like voluntary medical male circumcision in Tanzania.
Intra Health International uses R and QGIS for geospatial analysis and Power BI for dashboards, prioritizing free and open-source tools.
Finnegan is exploring Large Language Models (LLMs) via GitHub Copilot to improve the development and maintenance of open-source health software.
Building trust through clear communication, demonstrating value, and creating solutions that benefit stakeholders is critical for data scientists in global health.

Chapters

0:00 Introduction to Human Stories in AI and Amy Finnegan's Background

Series highlights AI experts' career journeys, sponsored by Lightning AI.
Amy Finnegan is Deputy Director of Data Science at Intra Health International.
She has over 10 years of experience in global health and data science across four continents.
Finnegan is also an adjunct faculty member at Duke University's Global Health Institute.

2:06 PhD Journey: Data Science Before It Was a Field

Finnegan earned her PhD from Duke University in 2016, initially in Political Science, then focusing on Sociology and Demography.
Her advisors mandated using secondary data and avoiding primary data collection to ensure timely graduation.
She learned skills now common in data science, such as merging datasets, quasi-experiments, and A/B testing, before 'data science' was a recognized field.
Her interdisciplinary program emphasized diverse approaches to the same questions, highlighting the common ground of regression and statistics.

5:34 Postdoctoral Research and the Value of Programming Skills

Completed a postdoc in Public Health at Duke Global Health Institute, transitioning to fieldwork and primary data collection.
Sought out a mentor who automated data collection and cleaning in R, recognizing the importance of these skills.
Received advice that a social scientist who can program stands out and can solve problems from any direction.
The postdoc focused on moving from secondary data to collecting primary data in public health settings.

11:49 Transition to Intra Health International and Building Data Systems

Worked as a Research Scholar at Duke, then transitioned to Intra Health International after a consulting opportunity arose.
The consulting role, initially a two-month project, became a two-month job interview, leading to her hiring as a Senior Data Scientist.
She emphasizes that job seekers should focus on what they can deliver, not just what they can gain from an opportunity.
Her current role involves building the 'Ready' system to unify data from various projects for executive and board reporting.

27:17 Project in Tanzania: Optimizing Voluntary Medical Male Circumcision

Addressed donor pressure to increase circumcisions with fewer resources.
Used high-resolution Facebook population density maps (1x1 km) and Demographic and Health Surveys (DHS) data to identify optimal locations for mobile testing and circumcision teams.
Integrated PEPFAR data on circumcisions performed and HIV prevalence to prioritize wards with higher HIV rates and unmet circumcision targets.
The project aimed to prevent HIV by increasing the proportion of the male population that is circumcised.

Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, StatQuest with Josh Starmer.