QBC LLC

Bryan Cook

AI Value Champion

Construction has a productivity problem and it isn't a work ethic problem. It's a visibility problem. Our project data lives across an ERP, a CRM, Google Drive and a handful of other places. By the time somebody stitches it into a picture you can actually decide from, the project has moved two weeks down the schedule and the picture has already changed.

We're a $24M multi-scope commercial subcontractor based in DeForest, Wisconsin. We started using AI in earnest about a year ago, not because we wanted to be early adopters, but because the gap between what we knew and what we could act on was getting expensive. Three uses have moved the needle for us. None of them are flashy. All of them are paying for themselves.

The big bet: a project productivity dashboard that finally tells us the truth.

We built a dashboard that pulls loaded man-hours, material costs, overages, unexpected expenses, change orders and many other metrics into a single view. It draws from our ERP, our CRM, Google Drive and the other places project information happens to live. Before this existed, that picture was assembled by a person, in a spreadsheet, on a Friday afternoon, two weeks late. Useful for the postmortem, useless for preventing losses.

We prototyped it with Claude. Once we had proof of concept, we handed the build to a development team to code it properly. That sequence matters. The AI was the cheap, fast tool for figuring out what the dashboard should be. Engineers built it to last. The result is a single pane of glass for every active project and the data is current enough that we can intervene before a margin erodes instead of after.

The team layer: a SMART goal app for internal projects.

We built a tool that helps our team turn fuzzy objectives into actual SMART goals for larger internal initiatives. The intent was simple. Lower the activation energy. If you make it easy for somebody to write a tight, accountable goal, more tight, accountable goals get written. We were tired of internal projects drifting because nobody had pinned them down properly at the start. The app does the pinning for you and the people closest to the work are more likely to use it than they are to wrestle with a spreadsheet template nobody loves.

The personal layer: daily AI on every executive's desk.

The least glamorous use case is also the one I would give up last. Every morning I get an AI-built briefing of my calendar, a debrief of my inbox flagging what actually needs my attention and a set of predrafted responses I can edit and send. Twenty minutes of my morning, every day, that used to be reactive triage. Now it's read, decide, move on.

The compounding effect across a year is real. More importantly, I show up to the rest of my day with a clearer head, which is a thing that doesn't appear on a spreadsheet but that you absolutely feel.

The pattern.

AI is not replacing judgment in our business. It's removing the friction between the judgment and the action. The dashboard doesn't tell us what to do about a struggling project. It just makes the struggle visible early enough to do something about it. The SMART goal app doesn't set the goal. It formats the thinking. The daily briefing doesn't decide. It queues the decisions cleanly.

For a midsized subcontractor, that is the right job for AI right now. We aren't trying to replace estimators or project managers. We're trying to give them better inputs, faster, with less of the busywork that used to eat the day. The ROI on that is real and it shows up in margin, in cycle time and in how much energy our people still have at 4 p.m.