As organisations move from AI that assists to AI that acts, agentic systems introduce new challenges around risk, accountability and control. Drawing on her work at Lloyds, Rajasree shares how agentic use cases such as investment AI and financial coaching are pushing platforms and observability beyond infrastructure level signals. She outlines how teams are evolving observability to understand agent behaviour, apply the right controls and coordinate across platform, data and governance functions, treating agentic AI as a shared enterprise ecosystem rather than a standalone capability.
• Extending observability to capture agent behaviour, decisions and outcomes beyond infrastructure metrics
• Establishing controls and ownership models that support safe operation of high risk agentic use cases
• Enabling enterprise scale agent adoption through collaboration, clarity and intentional constraint