Method

Plans built by an engine, checked by a validator

Most apps hand you a static PDF or a chatbot guess. Kalyra runs your answers through a deterministic programming pipeline — the same rules every time, auditable, and testable.

1. Screen first (Kalyra original activity-readiness screening)

Before anything is programmed, a standard activity-readiness questionnaire (adapted from widely used exercise-screening principles) plus pregnancy/postpartum status and injury flags are evaluated by a rule engine. Cardiac or chest-pain red flags route you to a gentle mobility track and a recommendation to speak with a qualified professional — we don't guess, and we never diagnose.

2. Constraints, not vibes

Your answers become an explicit set of constraints — e.g. second/third trimester means no back-lying, no jumping, no loaded rotation; early postpartum means crunch-pattern work waits while deep core strength rebuilds; a knee flag excludes deep knee flexion. Every exercise in our controlled library carries its own contraindication list.

3. Evidence-based programming

The engine selects from balanced movement patterns (squat, hinge, lunge, push, pull, core, cardio) following established strength-training principles: progressive overload across a 4-week mesocycle, rep ranges matched to goal, rest periods matched to effort, and a recovery-adjusted final week — a true deload for strength work, while mobility and gentle plans follow their own flow-based progression without one. Warm-ups and cool-downs are part of every session, not an afterthought.

4. Validate every single plan

Every generated program passes a validator that asserts the invariants: no contraindicated exercise slipped through, volume caps per muscle, session duration within tolerance, progression actually progresses, recovery weeks are actually lighter where a deload applies. Our test suite generates and validates thousands of programs across pregnancy trimesters, postpartum stages, injuries, equipment sets and schedules.

Where AI fits (and where it doesn't)

An AI language model will never invent your workout here. Programming is deterministic. AI's role in Kalyra — now and going forward — is communication: explaining why your plan is built the way it is, encouragement, and adapting tone. Prescriptions come from the engine; the engine is checked by the validator.

References we build on

• ACSM's Guidelines for Exercise Testing and Prescription — FITT-VP principles, pre-participation screening
• Established exercise-screening principles (general population readiness questions) — screening basis. Kalyra's check-in is original wording, not a licensed PAR-Q+ instrument.
• WHO physical activity guidelines — weekly activity targets
• ACOG guidance on physical activity during pregnancy and postpartum
• Progressive overload & mesocycle periodization as practiced in strength & conditioning (NSCA)

Note: Kalyra is a fitness product, not a medical device. Always consult qualified professionals about your specific health situation.