M Many msmes face challenges in making business decisions due to limited access toinformation and analytic technology. the research was conducted at sayur express, which sells retail vegetables online. the implementation of a business intelligencesystem is expected to help in making the right and fast business decisions so that sayurexpress can grow and develop. this study aims to develop vegetable product clustersbased on the k-means machine learning algorithm, compare the effectiveness of themean distance to centroid (wdc) value with the davies bouldin index (dbi) value forvalidating the number of vegetable product clusters, build a business intelligencedashboard, and integrate it with information systems for production, order, inventory, and sales analysis. this research uses a quantitative approach with data analytic methods. the clustering process is done with the k-means clustering algorithm, while datavisualization uses power bi. the data used comes from online vegetable salestransactions during may 2024. dbi is more effective for determining the optimalnumber of clusters, which is at an index value of 0.052. the optimal number of clustersis k = 2. the consumer segment in cluster 0 is 28% while in cluster 1 is 72%. the twoconsumer segments based on the clustering results are cluster 0 (28%) and cluster 1(72%). the developed dashboard helps analyze things like daily production tends toexceed orders and orders fluctuate inconsistently. the dashboard also helps analyzedaily sales fluctuations, favorite and non-favorite products. integration of businessintelligence with purchasing, production, inventory and sales information systemsresults in faster sales performance monitoring, identification of leading and non-leadingproducts, operational and stock optimization, and daily trend analysis.