A deterministic engine, grounded in exercise science.

The principles behind the engine.

• Muscle-specific, not whole-body

  Recovery is estimated per muscle group, because a single overall score can’t tell you that your legs are ready while your back still isn’t.

• Deterministic and transparent

  The same training inputs always produce the same recovery outputs, and contributors show what shaped each estimate.

• Uncertainty shown honestly

  Estimates are presented qualitatively with their uncertainty, rather than as false-precision numbers.

• Wearables enhance, never gate

  Optional signals from WHOOP or Apple Health can refine the picture, but the engine works from logged training alone.

• Guidance, not diagnosis

  Rekuvr informs training decisions. It does not diagnose injuries or provide medical advice.

• Built to be validated

  The engine is tested against a fixed suite of expected outputs, so changes are measured rather than assumed.