Brittany Irwin, AI Engineering Manager, NFI Industries: Why Your AI Rollout Isn't Really About AI

On this episode of People Solve Problems, host Jamie Flinchbaugh welcomes Brittany Irwin, Applications and AI Engineering Manager at NFI Industries. With an industrial engineering degree from the University of Pittsburgh’s Swanson School of Engineering and a career spanning large-scale third-party logistics and fast-moving B2B SaaS startups, Brittany offers a grounded view of how people, processes, and technology fit together. The conversation centers on a theme she has presented publicly, including at the Lehigh CSCRL Spring Symposium: how organizations move AI from hype to habit.

Brittany makes a case that runs counter to a common assumption. Most AI and automation efforts in logistics do not stall because the technology falls short, she explains, but because the business was never ready for it. The industrial engineering instinct to find waste and standardize it is the same discipline AI demands, since a tool can only return a reliable output when it is given a reliable input. So before moving any process toward automation, Brittany asks a pointed set of questions: where the process begins and ends, what the top exceptions are and why they happen, and whether any of it is actually written down.

That last question leads to what Brittany finds most underestimated, which is language itself. Being on time, she notes, can mean leaving the dock to one team and reaching the customer to another, and AI cannot reconcile a definition that people have never agreed on. This is why she puts such weight on a single source of truth. Consistent data definitions, documented exceptions, and shared context are what let a team build trust, reduce rework, and prepare the ground before automation is switched on.

Just as important, and far less often discussed, is the politics of change. Brittany asks two questions of every process: who gains power if it is standardized or automated, and who loses it. The first reveals a natural champion. The second reveals the person most likely to slow things down, and she reads withheld information and dragging feet as signs of exactly that. When the resistance outweighs what a project can overcome, she takes it as a cue to redirect her energy elsewhere.

Brittany is also reassuring about the fear underneath that resistance. AI, she insists, belongs in the second chair, not the first. It will suggest a course confidently and sometimes be confidently wrong, which is why a human must stay accountable for every decision, approval, and exce