AI built for nuclear engineering and medical physics.

What we are building

An instruction and reasoning platform built specifically for these fields. It teaches, assesses, and adapts to the learner, and it is grounded in the physics, the numerical methods, and the professional standards of the domain rather than in a general model's approximation of them.

The architecture is domain-independent; the specialized content is what makes it valuable. We begin in nuclear engineering and medical physics.

How we keep it correct

• Validated foundations. The system rests on computational methods validated against established reference codes and analytic benchmarks.
• Open methods. Our research is developed for peer review and for presentation at ANS and AAPM meetings, so the methods behind the product are open to scrutiny.
• One body of work. Every method produces a paper, a teaching module, and a component of the product. The research and the teaching are the same work seen twice.
• Human judgment. The system supports scientists, engineers, and clinicians. It does not replace their review.

Coming soon

The tutor opens to its first users with the first course cohort in spring 2027, and individual subscriptions follow later that year.