<p>Despite significant technological advancements, many business users lack the knowledge and expertise to effectively utilize business intelligence tools and machine learning techniques. This study aims to identify the optimal settings for developing predictive models tailored to SME business users. It focuses on determining the most robust machine learning model for revenue prediction and proposes deploying the model through an interactive financial dashboard, providing enhanced monitoring and decision-making capabilities. The Cross-Industry Standard Process for Data Mining (CRISP-DM) framework was enhanced by refining data understanding, preparation, model development and deployment using tools such as Power BI, JMP, and Python. For model development, machine learning algorithms such as XGBoost, random forest regression, ANN, and two hybrid models were tested. Additionally, an effective financial dashboard monitoring system was developed which allows stakeholders to monitor the company’s revenue, costs, and financial health in a user-friendly manner to support informed decision-making. The findings show that by utilizing the enhanced CRISP-DM framework, optimal settings for data understanding, preparation, and model development were obtained. The “K-prototype + RFR” hybrid model exhibited superior performance with <i>R</i>² at 94% in revenue prediction, showing resilience to overfitting and effectively handling data noise. The developed financial dashboard monitoring system proved to be an intuitive and powerful tool that can assist stakeholders in efficiently tracking revenue and financial health. This study provides valuable insights into the application of advanced machine learning techniques for financial prediction and demonstrates the practical benefits of integrating predictive model results with user-friendly dashboard systems.</p>