Multi-Agent Reinforcement Learning for Autonomous IT Transformation: Securing DevOps Pipelines with Explainable AI and Cryptographically Agile Workflows
International Research Journal of Modernization in Engineering Technology and Science · 2025
This work proposes a framework that combines MARL to drive autonomous IT transformation by securing DevOps pipelines through integration with Explainable AI and cryptographically agile workflows.Our approach, powered by real-world data sets from public repositories and industry-standard sources, demonstrates how co-operative agents dynamically adapt and harden pipeline security to maintain continuous compliance with the evolving nature of cryptographic standards by offering transparent decision-making processes for the same.Human-centered design in the system allows for interpretability, enabling IT professionals to understand the instituted security mechanisms better.Extensive simulations and real-world validations indicate dramatic gains that may enable the path to come up with robust, adaptable, and secure DevOps environments against current cybersecurity challenges in threat detection, response time, and system resilience.