The creation of mobile app has emerged as an integral aspect of modern organizations in a wide range of fields. A greater percentage of mobile app development (MAD) industries are turning to agile practices to increase app quality and shorten development times. Agile-based MAD relies heavily on efficient requirements gathering from stakeholders, and it has a significant impact on whether or not a mobile app is successful. Conventional methods of gathering requirements are typically labor-intensive and error-prone, prolonging the building process, driving up expenses and may struggle to maintain the track of frequent variations that characterize agile-based MAD. Machine learning (ML) approaches are becoming increasingly prevalent in gathering requirements as a means of automating the monotonous task of requirement engineering. The ability of ML approaches to process massive volumes of data, identify trends, and arrive at precise forecasts has the potential for significant advancement in gathering of requirements. This study intends to offer a methodological proposal for automating requirements gathering through the practice of a ML-based interactive bot, which we have named as requirements gathering bot (RGBot). The RGBot streamlines MAD process by acquiring requirements and optimizing customer experiences to meet stakeholders' expectations.