A three-session training built for the managers and directors who run the day-to-day at commercial GCs, MEP contractors, and home services firms. Real workflows. Real files. No theory.
Your PM is already pasting drawings into ChatGPT. Your controller is asking it to explain a contract clause. Your dispatcher is using it to draft customer emails. They're doing it without rules, without protocols, and without knowing what they shouldn't share.
This training fixes that. Two hours per session, three sessions, built for trades managers and directors. Hands-on, with real files and real workflows. No survey of the field; no buzzwords; no slides about the future of work.
Each session is two hours, hybrid (some hands-on exercises, some demo), in a conference room or on video. Bring a laptop. Bring a real file from your work. Sessions repeat monthly; cohorts cap at twelve attendees so the room stays small enough to do actual work.
Geoff Woods' CRIT methodology (Context, Role, Interview, Task) for writing prompts that actually work. The five biggest myths about what AI does and doesn't do, set straight in plain language. The protocols your team needs in writing before someone pastes a client contract or a payroll spreadsheet into a public AI tool.
We pick up where Session 1 left off. Real work on your own files. Inbox summaries that surface what actually needs your attention. Contract analysis that finds the clauses your team would miss on a Tuesday afternoon. Financial reporting that turns a spreadsheet into something you can show your owner without redoing the formatting.
The trades-specific session. Drawing packages and construction-document sets read with AI assistance. Multi-step workflows that chain together; the prompts your PMs and superintendents can save and reuse. The point isn't to replace the work; it's to remove the parts of the work nobody wanted to do anyway.
Before we teach your team about AI, it's worth knowing what we do when we build with it. Three recent engagements:
A commercial GC's COO was running daily inbox triage between sips of coffee. We built two Make.com workflows on the Claude API. One briefing in the morning, one in the afternoon, both delivered before he opened Outlook. Automation is easy to demo and hard to operate; this one's been running every day in production.
A COO was running a $30M revenue plan off CMIC reports he couldn't read. We sat in the requirements call, translated his verbal direction into 13 KPIs with the actual DAX formulas, and handed the developer a spec package complete enough to estimate from. No discovery meetings. No business-rule re-litigation.
A commercial GC was spending across hundreds of subcontractors with no tiered relationship strategy. We pulled the data and built the picture: top 20 vendors at 61% of spend, 53% of vendors used on a single job, and $1.15M in landscaping concentrated with one sub. The kind of exposure that sits in a founder's head until it doesn't.
The owner buys the training. The team takes it. PMs, operations managers, controllers, the dispatcher's boss. People who run real days at real companies, who don't need a lecture on what AI is, who do need to know what their team can safely do with it tomorrow morning.
If you're the owner and you want to attend the first session yourself before sending your team, that's encouraged. The room is built for operators.
Anthony Brandt runs the sessions. Background in construction and contract furniture; he's walked job sites and managed installations, and he runs Iron River as a firm that builds the systems most consultancies only recommend.
Sessions cap at twelve attendees. You bring a laptop and a real file from your work. We do the work together. You leave with prompts you've already used, protocols you've already drafted, and an understanding of what your team should and shouldn't be doing with AI starting that week.
If you finish Session 1 and decide the rest isn't for you, we'll refund Sessions 2 and 3. We'd rather know we weren't the right fit than keep the money.
Session 1 dates are posted on the registration page. If you want to talk first, book a call with Anthony and ask whatever you need to ask before committing.
Not sure if this is the right starting point? Take the AI Readiness Assessment instead.
