Based in the Dallas / Fort Worth area.
Modern AI is poorly aligned with the way science works. The architectures that dominate the field obscure their own reasoning and set aside centuries of hard-won scientific knowledge, demanding blind trust from researchers, clinicians, and patients.
In high-stakes medical tasks, that trade-off is untenable. A result that cannot be interrogated is a result that should not be trusted. Both patients and providers deserve to know why a decision was made.
Many of the assumptions underlying scientific and medical practice already have a precise, quantifiable form. At Weyl Labs we design novel architectures that exploit these forms directly, giving transparency to their mechanisms and confidence in their interpretations. Our models show where these scientific assumptions hold and where they break down, distinguishing the science from the statistics. A medical researcher can interpret the results alongside the scientific reasoning that produced them, judging each one against known science rather than in spite of it. Above all, these architectures are tools for researchers, designed to extend their judgment rather than override it.