AN ANSWER YOU CAN’T INTERROGATE IS AN ANSWER YOU CAN’T TRUST.

A black box has no place between a surgeon and an eye. Ask RLES AI why — why this hinge direction, why this optical zone, why this eye was flagged — and it shows the rule, the formula, the threshold and the paper.

EVERY NUMBER HAS A PEDIGREE

Ask why the PTA ceiling tightened on a steep cornea, why the hinge went temporal, why the borderline index count came to two and not four — and the answer names the rule, shows the arithmetic and cites the literature it stands on. Recommendations use the language of clinical decision support throughout: "recommended", "should be evaluated", "the decision rests with the surgeon" — never verdicts about patients.

A CONSULTANT IN THE POCKET, A TEACHER BY HABIT

• Residents can ask fundamentals — how ART max is derived, what the Alpins difference vector means — and get structured, sourced teaching
• Specialists can challenge a recommendation and see exactly which input drove it, then correct the input rather than distrust the output
• Feature questions on the line are answered with a short explanation plus the relevant guide PDF, automatically

WHY EXPLAINABILITY IS A SAFETY FEATURE

A surgeon who can interrogate a recommendation catches the wrong input that produced it. Explainability isn't documentation garnish — it is the error-correction loop that keeps a decision-support system honest, and it is why every RLES AI report shows its reasoning instead of just its conclusions.

FREQUENTLY ASKED QUESTIONS