Corporate financial distress arises from internal and external factors such as high debt levels and poor financial performance and causes significant challenges for businesses. Assessing the financial stress is a major task in the corporate houses. Previous researches on predicting corporate financial distress used several machine learning like decision trees or support vector machine that reveals several gaps. To overcome these issues, in this work, a detailed analysis on data of 422 companies with 3772 records and 84 features has been performed to indicate the stress level of the companies. To further investigate the research, financial distress prediction is done using the integration of ensemble learning techniques and feature reduction. The study employs multiple machine learning models, including random forest and gradient boosting, and combines them with feature selection methods to efficiently handle high-dimensional data. This approach not only improves the classification accuracy of distressed and non-distressed firms but also optimizes the computational efficiency of the models, offering a more robust solution for real-time financial monitoring and risk mitigation. The proposed model has the highest accuracy of 85.2%. This outcome indicates that the random forest model is highly effective in predicting financial stress levels in corporate houses.