IndiFitAI: An Intelligent System for Personalized Gym Workout Recommendations Using Machine Learning

While digital fitness applications have expanded access to exercise guidance, the abundance of available routines often leaves users struggling to find programs matching their specific physical traits and objectives. Conventional platforms typically depend on rigid templates or basic rules, frequently producing generalized and potentially unsafe training plans. To resolve this, we introduce FitAI, an intelligent, domain-tailored recommendation framework. By evaluating metrics like height, weight, gender, BMI, and fitness targets, FitAI accurately determines the ideal exercise categories for each user. The architecture leverages an ensemble machine learning methodology—featuring Random Forest, Decision Tree, K-Nearest Neighbors (KNN), and XGBoost—paired with BMI-based feature engineering for heightened precision. Tested against a curated Body Measurements Dataset containing 3,000 samples, the ensemble strategy proved highly effective. Random Forest attained a peak predictive accuracy of 89.7%, surpassing both KNN (88.2%) and XGBoost (83.1%). These findings confirm that FitAI provides secure, adaptive, and highly customized fitness paths, representing a substantial improvement over standard rule-based alternatives.