This study presents the implementation of the K-Means algorithm model, revealing novel insights into risk categorization in the delivery process. Two distinct clusters were identified: Cluster 1 (C0) indicating high risk, comprising 53 data points out of a dataset of 360, and Cluster 2 (C1) indicating low risk, encompassing 307 data points from the same dataset. Analysis conducted using RapidMiner Studio corroborated these findings, further delineating the cluster membership: C0 with 53 data points and C1 with 307 data points. Each cluster was characterized by optimal centroid values, recorded at 131.717 & 385.075 for C0, and 119.932 & 111.414 for C1. The model's effectiveness was assessed using the Davies-Bouldin Index, yielding a value of 0.626.