Designed and delivered a multi-agent AI system to automate complex operational workflows within a global data and insights environment. The system coordinated multiple agents, tools and data sources to execute tasks end-to-end, removing dependency on manual intervention and enabling consistent, scalable execution. The architecture combined structured orchestration with model-driven decisioning, allowing agents to interpret inputs, call tools, validate outputs and pass context between steps in a controlled and reliable manner. This was not a prototype. It was deployed into a live operational environment, supporting real workflows used by internal teams.