American cyclist Kristen Faulkner revealed she spent over 10 hours daily for two months programming with her personal biometric data to create an AI-driven training model.
The model processed her heart rate, sleep, weight, power output, and menstrual cycle phases, comparing them against 4,400 hours of her training history spanning nine years.
Faulkner stated she undertook the project because existing research lacked sufficient focus on elite female athletes’ performance needs.
She said the AI-generated insights proved invaluable, directly contributing to her winning three gold medals at the recent Pan American Championships in individual and team track events, plus the individual time trial on the road.
Faulkner, who won Olympic gold in the women’s road race and team pursuit at Paris 2024, had limited racing in the prior season due to a concussion sustained in a December 2024 training crash.
She began cycling in 2017 after moving to New York to work in private equity following her graduation from Harvard, initially joining women’s training sessions in Central Park.
The athlete emphasized that no single app provided a complete picture of her physiological data, prompting her to synthesize the information herself using AI tools.
What specific data did Faulkner’s AI model analyze?
The model analyzed her heart rate, sleep, weight, power output, and menstrual cycle phases.
How long had Faulkner been collecting biometric data before building the AI model?
She had been collecting biometric data for nine years prior to developing the model.