Many enterprise AI programmes begin in the wrong place. They start with interesting use cases, technology demonstrations or broad process maps, then struggle to connect those ideas to measurable business impact.

This session introduces a more practical approach: start where work is concentrated, understand how that work is actually performed and identify the parts of the workflow where AI can create meaningful operational value.

Drawing on a workforce-to-workflow discovery model, the session explores how organisations can move from broad AI ambition to targeted automation opportunities by mapping functions, decomposing roles, separating repeatable activity from judgement-heavy work and creating a clear view of where AI can realistically take ownership of subprocesses.

We’ll discuss:

Why chasing isolated AI use cases often leads to low-ROI pilots
How to identify the functions and workflows with the greatest automation potential
Why high headcount alone is not enough, and how repeatability and process stability shape AI value
How role decomposition helps reveal where time is actually spent
How to separate AI-owned subprocesses from human-owned judgement, negotiation and exception handling
How an AI density heatmap can turn discovery work into a board-level prioritisation conversation
What leaders should measure, including time saved, labour intensity reduced, cycle time improved and workflow ownership clarified

Rather than treating AI as a technology overlay, this session focuses on the operating model beneath it: where work sits, how it moves, who owns it and which parts are ready to be redesigned or automated. Attendees will leave with a clearer understanding of how to assess workflow opportunities, prioritise AI investment and move from experimentation to measurable operational impact.

Don't Start with AI Use Cases. Start with the Workflow

Speakers

Rory Clarke