Most fitness apps provide generalised diets and workout plans which might not be effective for the user. By tackling this problem using machine learning models, Onyx provides a way to personalize workout and diet suggestions by taking into consideration the physical state of the user with increased accuracy. Personal data like the amount of sleep, tiredness and soreness of muscle and recovery time is taken from the user and workouts are generated based on them. An example of this is how a user might be tired after a workout and his/her muscles may become sore. In such a scenario, ordinary fitness apps do not take that into account and provide workouts for the next day based on just the target that the user wants to achieve. Onyx has an edge over ordinary fitness apps by optimizing workouts based on the user's body condition. If the user is unable to progressively overload in the case of muscle soreness, Onyx provides workouts which do not increase strain on the muscles but at the same time does not compromise on the quality of workouts for instance increasing the reps and reducing weights.