Enterprise Agentic AI
Shivakumar R Goniwada · Apress eBooks · 2026
This chapter positions agentic AI as the next operational layer of enterprise intelligence, moving beyond AI that merely generates answers to AI that can perceive signals, reason under constraint, retrieve evidence, coordinate across multiple agents, and execute actions through governed tools. It frames agency as bounded autonomy, so agents can propose, plan, and prepare changes, but high-stakes commitments must remain accountable through policy, approvals, and system-of-record verification. Using the human brain as a design metaphor, the chapter introduces the core components of agents such as perception, cognition, memory, world models, reward, and action and shows how these capabilities are work together to produce reliable outcomes rather than fluent narratives.