Part 5 — AI Agents Across Business Functions
Overview
Understanding agentic AI in the abstract is one thing. Deploying it inside a real organisation — with its existing workflows, its accumulated technical debt, its specific regulatory obligations, and its particular mix of people — is another entirely.
Part 5 moves from principles to practice. It examines how agentic AI actually behaves when it encounters real enterprises: what production deployments have revealed that pre-deployment testing could not, how organisations should sequence their journey through increasing levels of autonomy, and what deployment looks like inside the specific functions — HR, Finance, IT, product development — where most of the early value is being realised.
The part opens with evidence before framework. Chapter 15 draws on real-world deployments to establish what actually works, what fails, and why — providing the empirical grounding that makes the structured framework of Chapter 16 more useful and more honest. Together they form the diagnostic backbone of the part: the maturity model is only credible when it is anchored in the production experience that preceded it.
Chapter 15 examines lessons from real-world frontier deployments — the failure modes, the success characteristics, and the question every organisation must answer before deploying consequentially. Chapter 16 provides the five-stage maturity framework that translates those lessons into a sequenced path from individual productivity tool to orchestrated multi-agent intelligence. Chapter 17 applies that framework to the three internal functions where agentic AI is being deployed most rapidly: HR, Finance, and IT. Chapter 18 examines how agents are changing the product development lifecycle — from research through to release. Chapter 19 addresses the structural question that organisations reach once they have more than one agent running: whether to centralise ownership or embed it in business units, and what each model implies for governance, talent, and speed.
Chapters in This Part
| Chapter | Title | Theme |
|---|---|---|
| 15 | When Agents Meet the Real World: Lessons from the Frontier | Production evidence |
| 16 | Staging Your Agentic Transformation: A Maturity Framework | Deployment sequencing |
| 17 | AI Colleagues: Deploying Agents Across HR, Finance, and IT | Internal function deployment |
| 18 | Accelerating Innovation: Agents in the Product Lifecycle | Product development |
| 19 | Where Does the Agent Live? Centralised vs. Embedded Models | Organisational design |
Chapters 15 and 16 are foundational and should be read first. Chapters 17, 18, and 19 can be read in any order depending on which functions or structural questions are most immediately relevant to your organisation.
Building agentic AI and wondering why alignment is harder than the technology? Get in touch