In response to the urgent need for coronavirus treatments, this research focuses on leveraging bioactivity data collection and processing for efficient drug discovery, employing computational methods to predict potential antiviral compounds. Exploratory data analysis was performed to identify patterns and trends, and various molecular descriptors were computed for the Mpro inhibitors in the descriptor dataset preparation step. A regression model was trained using the developed random forest algorithm to predict the bioactivity of new compounds against the standard value and compared with other regression models. Additionally, a web-based application was created using the random forest model to allow users to obtain predicted bioactivity values against Mpro by providing information about molecular structures. Machine learning-based bioactivity prediction offers an intriguing plan for drug discovery, and the proposed work provides a comprehensive workflow for COVID- 19 drug discovery. The web application provides a user-friendly interface for drug discovery researchers to evaluate the potential of compounds against the Mpro target quickly.