An AI proposal can look excellent in a demo and still find no place in operations. To bring the two together, someone has to understand the problem, talk to the people who do the work and follow the application all the way to delivery. That bridge between artificial intelligence and work can sit in different roles.
Start with whoever knows the exceptions
People who know a process notice details missing from the formal description: the request that needs different handling, the information that arrives incomplete, the approval that changes case by case. An operations professional knows, for example, that two similar requests need different documents. Whoever configures the application needs that distinction so the deliverable is not consistent on the surface and wrong in use.
The partnership can start with one activity observed end to end. Find where the work loops back, waits or depends on interpretation, and pick a result small enough to test together.
Judge the contribution by evidence of work
Ask the person to show the starting situation, the selection criteria, the first deliverable, the corrections, what stayed manual and what did not work. That reveals diagnosis, collaboration and review skills. A list of familiar tools adds to the conversation, but it does not replace this evidence.
To develop this role, make room for people from operations to take part in the tests. If all the responsibility sits with whoever knows the tool, knowledge of the work arrives too late. Leadership also decides who decides whether to continue and who looks after the application after the experiment.
The goal is a team able to recognize opportunities, experiment with judgment and explain what it learned — a capability that stays when tools and job titles change. That is the heart of AI Thinking by Dani Hayoshi. See also AI for leaders.
Dani Hayoshi is the author of AI Thinking. Her work connects artificial intelligence, education and entrepreneurship. About Dani.