When financial difficulties are identified early, stakeholders have the opportunity to limit losses, make informed choices, and act promptly to prevent unfavorable outcomes. New cutting-edge methods, mainly using artificial learning and machine learning, are especially beneficial in making appropriate predictions and thereby help in decision-making. Therefore, this study aims to employ machine learning algorithms, enhanced by feature engineering, to predict the financial distress of Indian companies. XGBoost and Random Forest was identified as the top-performing algorithm across all scenarios, including the entire sample, feature selection, and using PCA. The Artificial Neural Network performed well only when PCA was applied. In contrast, K-Nearest Neighbors and Support Vector Machine exhibited strong performance using the entire sample and feature selection, but their performance declined when PCA was employed. The proposed models demonstrate robust predictive performance, thereby providing valuable insights into financial distress patterns and thereby helping various stakeholders make informed decisions.