<![CDATA[The rapid advancement of digital technologies and volatile market dynamics have challenged mining companies to optimize supply chain management (SCM) processes. PT Sentosa Laju Sejahtera (SLS) experiences inefficiencies in procurement and inventory, causing decision latency and increased operational costs. This study aims to explore how integrating Machine Learning, specifically the Random Forest algorithm, into the company’s Enterprise Resource Planning (ERP) system can enhance predictive capabilities, reduce costs, and strengthen operational resilience. A qualitative case study approach was employed, combining in-depth interviews with key managerial stakeholders, participatory observations of operational workflows, and analysis of 63,818 historical Purchase Order documents. Triangulation techniques ensured data credibility and reliability. Findings reveal that Random Forest integration transforms decision-making from reactive to data-driven, reducing procurement cycle times from seven days to 2–3 days and lowering Total Cost of Ownership by 15% through fewer emergency orders. Feature Importance metrics enabled management to identify key inefficiency drivers, improving strategic interventions. Organizational readiness and data governance were identified as critical factors for successful implementation. In conclusion, the integration of predictive analytics into ERP systems provides significant operational and strategic benefits, enhancing efficiency, agility, and competitiveness. The study contributes both empirically to the literature on digital business transformation and practically as a model for multi-site mining enterprises seeking data-driven SCM optimization.]]>
Copyrights © 2026
