Rickman Trade: a practical AI implementation day

A practical implementation day is a different exercise from an introduction to AI. The room already accepts that the technology works. The open question is which of their own processes are worth pointing it at, and in what order.

How the day is structured

We spend the first part mapping how work actually moves through the company, as opposed to how the process documentation says it moves. Those two descriptions are rarely the same, and the gap between them is usually where the wasted hours are. From there we work down to a shortlist of candidates, sorted by how often the task repeats and how costly a mediocre result would be.

What comes out of it

• ✳A written map of where the hours currently go
• ✳A shortlist of automation candidates, ranked rather than brainstormed
• ✳At least one candidate built far enough to judge whether it is worth continuing
• ✳A clear statement of which tasks should stay entirely with people

The last point matters more than it sounds. A day that only produces enthusiasm leaves a team trying to apply AI to work it is bad at, which is the fastest route to concluding the whole thing was overhyped.

What it is worth

The team now saves roughly 20 hours a week. Most of it came from the same place it usually does: recurring reporting and analysis that was being assembled by hand. Sales results being read and summarised manually. Reseller performance data being pulled together for review. Findings from site visits being written up into the same email structure every time.

Work of that shape is the best first target in almost any company. It runs constantly, the input is already structured, and a first draft that needs editing is still a large improvement on starting from an empty document.