Sick Risk: From Messy Data to an Illness Risk Model
Details
How do you turn noisy, inconsistent public health data into something genuinely informative?
Aaron Hall with Source Allies walks us through the development of Sick Risk, a system that computes illness risk scores by zip code using CDC wastewater surveillance data. He will focus on extracting a stable, interpretable signal from messy real-world data, and the tradeoffs involved in doing so.
Along the way, he will explore the Python-based, cloud-hosted architecture behind the application.
Finally, he taks a look at how modern AI coding assistants accelerated development, without drifting into "AI slop."
Hosted at Source Allies, Tuesday, July 28th 5:30p-7pm. In-person or streamed online.
Speaker Bio: Dr. Aaron Hall
GenAI & Data Science Consulting at Source Allies.
Aaron Hall is a hands-on AI engineer building real-world systems at the edge of modern machine learning. As a consultant at Source Allies, he ships production-ready solutions using generative AI and retrieval-augmented generation (RAG), focusing on what actually works beyond the demos. Previously, he was a Data Scientist at Los Alamos National Laboratory, applying machine learning at scale across high-performance computing environments and contributing to open-source ML infrastructure. With a PhD in Physics and experience spanning national labs, research, and full-stack development, Aaron brings a rare blend of technical depth and pragmatic engineering, helping teams turn powerful models into reliable, high-impact products.