I help small and mid-size operations find the constraint, fix the process, and deploy AI their own team runs.
The big strategy firms will tell you the same thing: most of the work in AI is people and process, not algorithms. The difference here is that the people-and-process part is not a workstream I hand to a junior team. It is my entire career. Fifteen years on production floors as an industrial engineer, plus the software skills to build the system myself, on your hardware, at a scale a local business can actually afford.
A practical field guide for the people who actually have to get things done. Ten chapters, twenty case studies, and every tool comes with the place where it breaks. If you want to know exactly how I will think about your operation before you ever book a call, it is all in here.
Each one stands alone, and each one earns the next. You are never locked into a roadmap you cannot leave.
I walk your operation, time the real process, and price the constraint. Then every possible AI use is ranked against that number, including the ones vendors are pitching you.
You get: a costed shortlist, one clear first project, and an honest list of things not worth automating.
Working software on your data and your hardware, scoped so a wrong answer is visible and cheap. Document extraction, drafting, routing, scheduling, quality flags. Judgment stays with your people.
You get: a system in production you judge working, not pitched. Measured against the constraint, in weeks.
The part that decides whether any of this survives: standard work, training, a named owner on your payroll, and a rollback plan so trying it was never a risk.
You get: a team that runs it after I leave. If they cannot, I did not finish the job.
Work flows, it piles up in exactly one place, and everything upstream of that pile is running faster than it needs to. Every engagement starts by finding that place.
Every figure below is from my resume, at the company named under it.
A value stream, then a line, then a plant, then a network. Two moves look like steps down and were not: in 2015 I stopped to learn to build software, and in 2020 I traded a slice of a huge network for every operation in a smaller company.
Need this as a document? The full resume lives at rochaine.com/resume, print-ready.
Twelve questions I ask on every floor, free. If you cannot check a box, that is the work to do first, and it is usually cheaper than the software.
No, and it works better if you are not. The strongest early wins are in warehouses, shops, fabricators, clinics, and back offices: places where people spend hours retyping, reading, and routing. If your operation has paper, spreadsheets, and a step everyone complains about, there is something here.
Every engagement is fixed-scope and quoted after the intro call, once I know the size of your operation and the problem. The assessment is priced so that a clear decision comes out of it, not a dependency on me. If the honest answer is that AI is not worth it for you yet, that is what the assessment will say.
The unglamorous kind that pays: pulling data out of documents, drafting the routine paperwork, routing and flagging work, catching quality problems early, scheduling against real capacity. Not a chatbot on your website. Judgment stays with your people; the machine takes the clerical load off them.
The big firms are right about one thing: most of the work in AI is people and process, not the model. The difference is that at a large firm that part becomes a slide deck and a junior team. Here it is one experienced engineer who does the floor work personally, builds the system himself, and hands your team the keys. You also do not pay for a tower of overhead.
I am based in Wellington, Florida and work on site across South Florida, from Palm Beach down through Broward and Miami-Dade. For the build and support phases, remote works fine anywhere.
Then I tell you, and you will have saved the cost of finding out the expensive way. Sometimes the fix is a layout change, a standard, or one more person at one station for part of the day. I am an industrial engineer first; the point is the throughput, not the technology.
What do you make or move, and where does it pile up? That is enough to start. I read these personally.