As 5G-Advanced increases network complexity, operations and maintenance (O&M) methods need to evolve to meet performance demands. Manual fault handling, service assurance, and network tuning are too slow and resource-intensive for today’s dynamic wireless environments. The RAN multi-agent system Catalyst introduces a fully digital engineering framework. It replaces manual intervention with intelligent agents and automation—streamlining RAN O&M across fault management, optimization, energy efficiency, and SLA assurance. The solution includes four AI-powered digital agents: FME Mate, Optim Spirit, Watt Spirit, and Bit Mentor. Each one is designed for a specific high-value use case. FME Mate handles real-time fault diagnosis using foundation models and live topology data. Optim Spirit uses digital twin simulation to drive continuous KPI-based self-optimization. Watt Spirit identifies optimal energy-saving policies that preserve performance while cutting consumption. Bit Mentor applies predictive analytics to deliver SLA-based provisioning and prioritized service for high-value users. Together, these agents operate within an end-to-end, lights-out O&M pipeline. Digital antennas and service processors enable real-time telemetry and network adjustments without site visits. The system continuously refines its recommendations using AI and live feedback, enabling rapid, closed-loop decision-making across the network. Results are already measurable. FME Mate reduces mean time to resolution (MTTR) by 30%. Optim Spirit closes 20% of performance tickets autonomously. Watt Spirit delivers 10% network-wide energy savings with no impact on user experience. Bit Mentor predicts SLA performance with 85% accuracy, enabling on-demand provisioning and differentiated service monetization. By automating complex workflows, this Catalyst transforms network operations into a predictive, efficient, and scalable process. Crucially, it gives CSPs the tools to manage 5G-A complexity while delivering stronger performance and lower costs.