An Explainable Multi-Agent Framework on SAP BTP for Intelligent Enterprise Decision Making

Abstract

The need for intelligent enterprise decision making has become a standard need for organizations that want to improve operational efficiency and a more agile business while providing transparent AI-driven services. Cloud-native enterprise platforms and AI technologies that are understandable make it possible to convert complicated enterprise information into certain and also understandable choices. This paper introduces an Explainable Multi-Agent Framework on SAP Business Technology Platform (SAP BTP), a step towards the intelligent enterprise decision-making. It combines autonomous software agents, SAP HANA Cloud, SAP Integration Suite, SAP AI Core and SAP Build Process Automation to work together to analyze enterprise information, make predictions that can be explained, govern and validate business rules, and automate the exchange of processes. Multi-agent collaboration is good for distributed reasoning and explain ability is good for giving reasoned explanation for all the business recommendations clearly showing to users why a certain recommendation is being made, thus enhancing reasoning and governance. Experimental results show that the proposed solution outperforms the best existing solution with regard to 96.8% Decision Consistency, 95.9% Workflow Automation Efficiency, 94.3% Explanation Fidelity, and 97.5% System Scalability benefits, while achieving a total reduction of 21.2% in Decision Response Time as follows: 2.4% in Decision Consistency, 2.7% in Workflow Automation Efficiency, 4.2% in Explanation Fidelity, and 3.3% in System Scalability. The proposed ‘framwork' is a scalable, trust proof and intelligent approach for next generation enterprise decision support for SAP BTP.