Podcast
From Models to Medicine: The Microbiome is Messier Than You Think
From Models to Medicine: The Microbiome is Messier Than You Think
Jenny Yang is the co-founder and CEO of Outpost Bio, where her team is working to make human microbiology computable. In this episode, she breaks down why bias in ML models is so easy to miss. High overall accuracy can hide terrible performance on specific subgroups, and in healthcare, that gap has consequences. She traces the problem upstream, from skewed training datasets to the way clinical definitions themselves carry historical bias, and explains the real trade-offs involved in trying to correct for it.
Jenny Yang is the co-founder and CEO of Outpost Bio, where her team is working to make human microbiology computable. In this episode, she breaks down why bias in ML models is so easy to miss. High overall accuracy can hide terrible performance on specific subgroups, and in healthcare, that gap has consequences. She traces the problem upstream, from skewed training datasets to the way clinical definitions themselves carry historical bias, and explains the real trade-offs involved in trying to correct for it.
Outpost Bio Co-Founder and CEO Jenny Yang joined Michelle Yi on KAMI Think Tank’s From Models to Medicine to talk about a challenge that sits at the heart of AI for healthcare: strong overall model performance can still hide serious bias—and those failures matter when models are used to make decisions about people.
The conversation covers:
Why high overall accuracy can mask poor model performance across specific patient subgroups
How bias can enter healthcare AI through both skewed training datasets and the clinical definitions used to label them
Why the microbiome creates an especially difficult modeling problem, with microbial communities varying dramatically from person to person
How Outpost’s Lab-in-the-Loop approach connects wet lab experimentation with AI to generate higher-quality, less biased biological data
Why rigorous external validation is one of the most important steps biotech teams can take before trusting a model in the real world
Outpost Bio Co-Founder and CEO Jenny Yang joined Michelle Yi on KAMI Think Tank’s From Models to Medicine to talk about a challenge that sits at the heart of AI for healthcare: strong overall model performance can still hide serious bias—and those failures matter when models are used to make decisions about people.
The conversation covers:
Why high overall accuracy can mask poor model performance across specific patient subgroups
How bias can enter healthcare AI through both skewed training datasets and the clinical definitions used to label them
Why the microbiome creates an especially difficult modeling problem, with microbial communities varying dramatically from person to person
How Outpost’s Lab-in-the-Loop approach connects wet lab experimentation with AI to generate higher-quality, less biased biological data
Why rigorous external validation is one of the most important steps biotech teams can take before trusting a model in the real world
Related articles
Legal Pages
©OUTPOST BIO 2026
BAKED BY THE SOURDOUGH
©OUTPOST BIO 2026
BAKED BY THE SOURDOUGH