Unlocking Unstructured Health Data with David Sontag from Layer Health

What if we could unlock the hidden potential of unstructured health data to improve patient outcomes? In this episode, I sit down with David Sontag, co-founder and CEO of Layer Health, to discuss the transformative role of AI in healthcare. David, an MIT professor (on leave) and leading machine learning researcher, delves into how Layer Health addresses one of healthcare’s most persistent challenges: extracting actionable insights from unstructured medical data. In our conversation, David explains how Layer Health’s AI platform automates complex chart review tasks, tackles data generalization issues across diverse healthcare systems, and overcomes challenges like bias and dataset shifts. We explore Layer Health’s groundbreaking use of large language models (LLMs), the importance of scalable AI solutions, and the integration of AI into clinical workflows. Join us to discover how Layer Health is reducing administrative burdens, improving data accessibility, and shaping the future of AI-powered healthcare with David Sontag.

Key Points:

• Hear about David's career journey from MIT professor to CEO of Layer Health.
• How Layer Health transforms chart reviews and enhances healthcare workflows.
• The role of large language models in solving the company's scalability problems.
• Learn about Layer Health's approach to benchmarking performance for institutions.
• Explore how the company navigates dataset shifts and ensures robust model performance.
• Discover Layer Health's strategies to identify and mitigate bias in clinical AI models.
• Find out about the challenges of implementing reasoning across diverse medical records.
• Why building trust through data transparency, auditing, and compliance are essential.
• David’s advice for AI startup leaders on balancing research with practical implementation.
• Layer Health's long-term vision for reshaping healthcare and improving patient outcomes.

Quotes:

“Our vision for Layer Health is to transform healthcare with artificial intelligence, really building upon all of the work that we've been doing over the past decade in the AI and health field and academic space.” — David Sontag

“What we realized very quickly is that where [Layer Health] would have the biggest impact was bringing the right information to the physician's fingertips at the right point in time.” — David Sontag

“We're using large language models to drive the abstraction of those clinical variables that we need for these either retrospective or prospective use cases.” — David Sontag

“Where I think we're going to see the biggest source of bias is likely going to be not along the traditional demographic-related quantities, but rather on more clinical quantities.” — David Sontag

“A lot of the friction that we currently see in healthcare, [Layer Health] is going to really take a big bite out of [it].” — David Sontag

Links:

David Sontag

David Sontag on LinkedIn

Layer Health

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.