Zihi represents a novel approach to training plan generation, what we call the continuously adaptive training plan. By gathering information from athletes about the time and resources they have to train we generate an initial training plan. As athletes complete workouts, they populate the plan with objective data from a wearable device (e.g., power from a power meter) and subjective information about how they are feeling (e.g., sleep quality, muscle soreness, recovery status, life stress, and personal response to a given training load). Using sports science-based AI algorithms, Zihi modifies the training plan to reflect the athlete’s progress with training, adjusting the intensity and volume of workouts. This approach offers advantages to both athletes and coaches. Think of Zihi as an “autopilot” for coaches, performing mundane tasks best managed by a computer and freeing up coaches to do what they do best: supporting and motivating athletes. By automating routine coaching tasks such as writing training plans and monitoring athletes' progress, Zihi allows coaches to take on additional clients and grow coaching revenue. With Zihi, coaches will ask athletes to enter their time and facilities available to train, and eventually, we plan to be able to API with an athlete's Outlook or Google calendar to import a schedule directly. Once an athlete has entered his or her availability, Zihi will produce a tailored training plan and update it as the athlete completes workouts and reports on other factors that moderate training performance and response. Coaches will have the ability to edit the plans Zihi produces using our Exception Rules tool, so there can be further customization based on the coach's experience or the coach's or athlete's preferences. We hope that an AI-based plan with input from a coach will allay athletes' concerns about Zihi.
