Developing Mobile App SupplierHub: Integrating Spatial EWM-SAW and Generative AI for MSME B2B Supply Chain Optimization

Abstract

MSMEs often face difficulties in identifying suitable business partners and establishing effective B2B communication, which can affect procurement efficiency and supply chain performance. This study develops "SupplierHub," a mobile-based decision support system designed to enhance partner selection accuracy while automating professional B2B communication. Adopting the Design Science Research Methodology (DSRM), the proposed system integrates Haversine-based spatial analysis with Simple Additive Weighting (SAW) method for multi-criteria decision-making. To ensure an objective evaluation, the Entropy Weight Method (EWM) is utilized to dynamically determine criterion weights based on real-world data distribution. Furthermore, a Large Language Model (LLM) is incorporated to automate the generation of professional B2B proposal drafts, bridging the communication gap in business collaborations. Experimental results with 60 partner candidates demonstrate that the proposed EWM-SAW model achieves a high Spearman correlation coefficient of 0.976 against the TOPSIS and VIKOR baselines, while significantly outperforming the subjective AHP-SAW model. Furthermore, scalability testing up to 10,000 records proves its computational superiority, executing over 75% faster than the matrix-heavy baselines. This ensures real-time operational efficiency without causing mobile client bottlenecks. Additionally, the generative AI integration is securely routed through a server-side proxy to safeguard sensitive MSME procurement data. The generated proposals passed multi disciplinary expert evaluation with a mean score exceeding 4.17 across all business metrics. Additionally, user evaluation yielded a System Usability Scale (SUS) score of 84.4. The results indicate that SupplierHub serves as an effective, highly scalable mobile decision-support tool for partner selection and B2B communication in MSME supply chains.