Sleep quality impacts physical and emotional wellbeing of humans. If sleep is distressed it leads to complications in day-to-day activities, resulting in stress, anxiety, and neurological disorders. The main objective of this study is to embrace proficiencies of machine learning (ML) algorithms to predict quality of sleep, and to detect disorders, if any. Specifically, current study is intended to meet three objectives. These objectives are: (a) select relevant data features, (b) identify best performing model, and (c) design robust ML-based sleep prediction tool. In this perspective, sleep dataset is collected and subjected to feature selection and classification tasks. Results from experimentations revealed six prominent features via Boruta algorithm. Furthermore, RBF-SVM, RF models predicted sleep disorders with enhanced accuracy. Also, performance of RBF-SVM was fine-tuned by feeding its predictions to polynomial kernel SVM. This model achieved an accuracy of 95.98%. Finally, a robust ML tool, SleepPred, was developed to predict sleep disorders.