The practice draws on mathematics and computer science without treating technology as the answer in advance. Small models and prototypes create evidence early, expose limits, and help decide whether a larger system is justified.