Explainable multi-agent AI systems for intelligent software engineering and business automation

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Generative Artificial Intelligence (GenAI) and Explainable Artificial Intelligence (XAI) are revolutionizing software programming and enterprise business automation in the 4th industrial revolution era. Current multi-agent systems, however, are constrained in being able to collaborate, in understanding how to reach consensus, in their knowledge sharing and, in their ability, to adapt to changes and therefore are not suitable for certain industrial critical applications. To address the above challenges, this paper introduces a cognitive swarm-based multi-agent framework with collaborative reasoning, collective cognitive memory, explainable decision intelligence, digital twin simulation and autonomous governance, which are integrated into an explainable cognitive swarm control framework XCognitive SwarmNet. Proposed framework mathematically models agent collaboration, trust, explainability, Knowledge sharing and Optimization of the flow of business tasks to support the complete Software Development Lifecycle and Intelligent Business process Automation. By using the powerful and representative software engineering and business process datasets, some experimental assessments are performed and the proposed mechanism is compared to six state-of-the-art multi-agent approaches. The proposed framework completes the task with 98.1% accuracy, reduces resource consumption by 22.8% relative to the selected baseline/reference configuration, achieves 97.5% in explainability fidelity, optimizes the workflow by 95.9% and optimizes the efficiency of agent collaboration by 94.7% with a 27.4% reduction in execution time. Overall, the results indicate that XCognitive SwarmNet provides a promising framework for transparent, adaptive, and trustworthy AI-assisted software engineering and business automation, while further industrial validation is required before claims of deployment in safety-critical or operational Industry 4.0 environments can be established.