Disruptions highlight machine learning (ML)-powered digital twins as a significant technology to ensure supply chain resilience (SCR). The authors acquired supportive data by conducting semi-structured interviews with 37 fast-moving consumer goods supply chain professionals. Using open, axial and selective coding approaches, the authors mapped and discovered the themes that constitute the essential elements of ML-enabled SCR. The findings of the research underscore four principal capabilities in which ML is poised to enhance the resilience of supply chains, namely (a) visibility, (b) supply chain analytics, (c) managing levels of inventory and (d) consumer behaviour. The research implications employ empirical data to demonstrate how integrating ML with digital twins into the supply chain enhances resilience and presents a comprehensive framework utilizing ML to improve SCR. The research draws upon the dynamic capability view theory.