3–5 workflows audited
Payroll, daily reports, change orders, intake, AP — whichever ones are bleeding. Real walkthroughs with your team, not surveys.
I spent the past year inside a $7M glazing sub — payroll, daily reports, change orders, compliance. I’m a developer by background, so when AI got good I started automating the back office instead of growing the team. Got the owner 20+ hours a week back. Now I do this for other SMB subs.
$3,5001 week1 founder
01The Audit
Before you spend real money on an automation build, you need to know exactly where your hours go and which workflows are worth touching first. That’s what the audit is.
Payroll, daily reports, change orders, intake, AP — whichever ones are bleeding. Real walkthroughs with your team, not surveys.
Each workflow gets an hours-saved estimate, build difficulty, and risk. You see the order of attack before signing anything bigger.
By the end of week one, at least one workflow is already saving hours. Proves the pattern before you commit further.
02Not for everyone
Three groups I won’t chase. If you’re any of these, we’ll both save a call.
different sale cycle, IT department, procurement
they have their own AI roadmaps
no recurring data drudgery to automate
03FAQ
That's exactly the point. Small ops teams have the most hours per person to reclaim. The audit tells us where.
Zapier breaks when the workflow needs judgment. The new tools (Claude, GPT) handle judgment. Different category.
Every output gets human sign-off in v1. Once we trust a workflow's accuracy we relax that. Same pattern as hiring a new admin.
Audit week 1 ends with a quick win already implemented. You'll see one workflow saving hours before you commit to a build.
04About
I’ve been the ops person, not a vendor with a deck. I know what dies on Friday afternoon when a foreman forgets a timesheet, and I know what it costs you when the GC kicks back a daily report three weeks running.
Technical depth predates the construction year — React/Next.js/Supabase/Python, AI tooling fluent. Built the automation stack inside a real $7M sub. Not learning on your dime. I work alone, fast. Small client, small contract — get a quick win, decide if you want more.