Leoson Hoay

United States · Singapore · Australia · leoson@uchicago.edu

I find the term "data science" clunky - but a lot of my work and research does involve analyzing and managing data, and applying the scientific method to engineer insights our chaotic world. My path in analytics has been neither straight nor narrow, and I have been walking the line between computer science, the social sciences, and the humanities for many years. I am most passionate about leveraging advances in technology and analytics to solve human problems, and bringing these fields closer together.

I am always up for a chat about causal inference, why the gradient keeps disappearing in your model, how researchers used text analysis to assert that Shakespeare was not a single author, and the metrics I used to rank the best cups of hot chocolate that you can find in Hyde Park.

I currently serve as a Data Scientist at Learning Collider.

Outside of my day job, I write a personal arts and science newsletter, Strictly Interdisciplinary, and co-host a youth talkshow and podcast on science and society, Very Clear Cut. I am also a Part 107 Licensed Commercial Drone Pilot in the US.

Current Location: Chicago, IL, USA


Experience

Data Scientist

Developing statistical analysis and machine learning models on partner data platforms to accelerate equity in education and housing.

Nov 2022 - Present

Technical Lead and Lead Data Scientist

Research Analyst and Data Steward
Data Analyst
Research Assistant

Lead the development of data science and analytic talent at the Health Lab. Provide analytic and data science support to various projects in the lab, working with partners such as the Illinois Department of Public Health and the Chicago Children's Advocacy Center in leveraging statistical models and developing analytical tools to evaluate public health policies and interventions. Maintain organization-wide data usage policy and data infrastructure, and reviews and develops new policies subject to HIPAA compliance, user feedback, and best practices in PII/PHI data governance.

Mar 2018 - Oct 2022

Data Engineer

Developed ETL workflows, supported data curation tasks, working mostly with Airflow and cloud platforms such as AWS and Google Cloud.

Apr 2019 - Sep 2019

Writing Tutor

Provided instructional and consultative writing support to the UofC student population in both academic and business writing.

Sep 2018 - Jan 2019

Interaction Designer

Prototyped in Unity and JavaScript for IoT and VR applications. UX design and research using mixed methods, lab studies, and device monitoring.

Feb 2017 - Jun 2017

Intern - Program Implementation and Evaluation

Review and evaluation of existing psychological and clinical program implementation. Building data infrastructure and data collection workflows, in-depth research and literature review of current clinical and rehabilitation practices in correctional settings.

Dec 2016 - Feb 2017
Dec 2016 - Feb 2017

Community Analyst

Aug 2015 - Nov 2015

Student Services Advisor

Aug 2014 - Nov 2014
Jul 2014 - Oct 2014

Education

Georgia Institute of Technology

Master of Science, Computer Science

Computer Vision · Machine Learning · Robotics

August 2020 - June 2022

University of Chicago

Master of Arts, Computational Social Science

Causal Inference · Machine Learning · NLP · Computational Linguistics · Databases

August 2017 - June 2019

Australian National University

Study Abroad Program

Science Communication · Anthropology · Philosophy

July 2014 - November 2014

National University of Singapore

Bachelor of Social Sciences (Honors)

Communications & New Media · Psychology

August 2012 - December 2016

Raffles Institution

Integrated Programme (Cambridge A Level)

Literature · Chemistry · Mathematics · Economics

January 2004 - December 2009

Awards & Certifications


Projects

  • COVID-19 Prediction using CT Scans
    A project that used machine learning models to predict COVID-19 diagnoses using CT Scan image data. This project was completed in collaboration with Fiona Fan as part of the requirements for the MS in Computer Science degree with the Georgia Institute of Technology.

  • Very Clear Cut
    Very Clear Cut (VCC) is a media platform that hosts monthly talkshows on Facebook Live with invited guests, on topics that are pertinent to the human condition - from environmental sustainability and career advancement, to mental health and education. In addition, we also host regular fireside chats on Clubhouse on these topics, where listeners and members of the public are invited to be a part of the conversation.

  • Transform911
    The University of Chicago Health Lab is gathering experts in health care, academia, government, emergency response and public safety—in collaboration with advocates and elected leaders via virtual roundtables and working groups—to examine America's 911 system.

  • A Primer to Folium: A Python Library for Simple Geospatial Visualizations
    A lesson I created as part of a Python workshop for beginners. This notebook explores a simple use case for Chloropleth maps in the Folium library, with a sample dataset of chess grandmasters. It also briefly visits the topics of web scraping and connecting to a database (MongoDB).

  • Visualizing the Mandelbrot Set in JavaScript and Canvas
    Using an escape algorithm, the fascinating Mandelbrot Set is rendered here using JavaScript and Canvas.

  • Exploring the Yelp Academic Dataset
    As part of a computer science course at the University of Chicago, my team and I explored the Yelp Academic Dataset in order to build a pair-wise restaurant recommendation system, as well to test the hypothesis of the benefits of spatial proximity between similar businesses. Various tools were used, such as Hadoop/MapReduce for parallel processing and gensim for text analysis.

  • Building a Dataset Search Engine with Django
    As part of a computer science course at the University of Chicago, my team and I developed a Django framework for building an open-source survey database, with search capabilities.

  • User Experience Portfolio
    Samples of user experience research and design work done in my college years.

Publications

Very Clear Cut