AI Adoption in Mid-Sized Firms: Start With the Workflow, Not the Model
AI pilots in mid-sized firms succeed when they start from a painful workflow, not a promising model. A practical way to choose your first three.
By Owen Hartley, Partner, operations and technology , 2 min read
The firms getting real value from AI this year did not start with a model. They started with a queue of work that annoyed everyone. AI pilots in mid-sized firms succeed when they start from a painful workflow, not a promising model. A practical way to choose your first three.
Pick the workflow, then the tool
Look for work that is frequent, rules-based and already measured: quote preparation, invoice matching, first-line customer questions. If nobody can say how long the task takes today, you will not be able to prove that AI made it faster. Spend the first two weeks measuring before anyone opens a vendor demo.
Design the human checkpoint
Every pilot needs a named person who reviews outputs and a clear rule for when the system must hand back to a human. That checkpoint is what turns an impressive demo into something a COO will sign off, and it is usually where the real productivity gain is found.
How we apply this with clients
In client work this usually becomes one priority on the quarterly plan, with a named owner and a measure the leadership team reviews every month. Our technology & ai partners start with a short diagnostic of where the business is today, then agree the smallest change that will make a visible difference within ninety days.
Key takeaways
- Start from a measured, frequent workflow, not a model.
- Name the human reviewer and the hand-back rule before launch.
- Judge the pilot on hours saved and error rate, not on enthusiasm.
Want to apply this in your business? Book a discovery call with a Maia Lafortezza partner.
Owen Hartley
Owen Hartley leads the firm's operations and technology practice. An industrial engineer by training, he spent twelve years running distribution and manufacturing operations before moving into advisory work. He redesigns workflows, operating rhythms and supply chains for logistics, manufacturing, healthcare and service businesses, and helps clients choose and implement technology, including practical AI, where it pays back. He is based in Milan.
More from Owen
Discussion