About Glia
In the nervous system, glial cells don't fire signals themselves — they support, nourish, and protect the neurons that do. That's the idea behind this site. Healthcare has no shortage of brilliant clinicians, researchers, and technologists. What it often lacks is the connective tissue: clear thinking about how data, AI, and interoperability can be made to actually work in practice. Glia exists to provide some of that support — helping the neurons of medicine do their best.
Why I write here
Healthcare AI and data are at an inflection point. The tools are genuinely powerful. The standards are maturing. And the gap between what's technically possible and what gets implemented well — safely, equitably, and usefully — remains frustratingly wide.
I write here because I've spent my career at that gap — bilingual in medicine and computer science, working both sides of it. I want clinicians to understand what AI can and can't do in their workflow. I want technologists to understand what clinicians actually need. There's no shortage of hype in this space. I'd rather offer something more useful.
Background
After training in medicine, I earned a PhD in biomedical informatics focused on health AI (PDF link) — before it became a buzzword. I'm a faculty member in Internal Medicine (Epidemiology) and Biomedical Informatics (adjunct) at the University of Utah. I conduct healthcare AI research across both the University of Utah Health and the Salt Lake City VA Medical Center. This blog reflects my own views.
My research lies at the intersection of artificial intelligence, data standardization, and answering complex healthcare questions in infectious diseases, epidemiology, oncology, mental health, and other medical specialties. My current research involves the use of AI for natural language processing (NLP).
Before returning to academia, I spent over 15 years in industry building healthcare AI and data infrastructure at scale. At 3M Healthcare (now called Solventum), I worked on clinical terminology, data standardization, and NLP — the invisible but essential foundation that makes AI in healthcare possible. At Truveta, I worked on data standardization using human-in-the-loop AI to harmonize real-world clinical data across health systems. Most recently, I led Nightingale Open Science incubated at the University of Chicago, where I led finance, cloud data and GPU computing strategy, and multimodal AI research — using images, waveforms, and videos to investigate what clinical AI can learn from diverse data types.
Along the way, I served as an elected U.S. representative on the SNOMED International Technical Committee and advised a U.S. Congressional Committee on improving interoperability between VA and DoD medical records. I'm a Fellow of the American Medical Informatics Association (FAMIA).
What I write about
This site is for clinicians, technologists, administrators, and policymakers who want to understand AI, data quality, interoperability, and patient safety — without the hype and without oversimplification. I write to be understood across disciplinary lines: a clinician shouldn't need a computer science background to evaluate an AI claim, and an engineer shouldn't need medical training to understand why data quality matters at the bedside.
When a topic can't be covered that way without losing depth, I write separate articles for different audiences on the same subject and note who each one is for. I occasionally consult with healthcare organizations and technology companies on these topics — from startups to large health systems — when I can be helpful. The articles here reflect the questions I get asked most, and the ones I think matter most.
Get in touch
If something here is useful to you, I'd like to know. If you have questions, disagreements, or something you'd like me to write about, I'd like to know that too. You can reach me through the contact form, or find more about me at:
LinkedIn | X / Twitter | ORCID | Google Scholar | University of Utah faculty page
Senthil K. Nachimuthu, MD, PhD, FAMIA.