DeepNeuro.ai is a research practice at the seam of AI safety, clinical AI, and neuroscience. We evaluate how foundation models fail at deployment scale — adversarial, clinical, decision-support — and ship the benchmarks, datasets, and production systems needed to deploy them safely.

It is led by Richard J. Young, PhD, Senior AI Research Scientist at UnitedHealth Group and Assistant Professor-in-Residence in Information Systems at the UNLV Lee Business School, with Alice Matthews, PhD, RDCS on clinical imaging. Recent work includes CardioEmbed, a clinical cardiology embedding model that reached 99.6% retrieval accuracy on cardiac-specific tasks (arXiv 2511.10930, November 2025), and EQUITRIAGE, a fairness audit of LLM-based emergency department triage.

The work has a north star: a measurably better world, built with AI we can measurably trust.

Updated September 2026