For industrial operators
I work with manufacturers, distribution centers, shipyards, terminals and the contractors who serve them. I find where time is lost and fix it, through better processes, AI tools and training.
Request a 30-minute callEvery shift produces two things. One is the work: the parts made, the orders shipped, the vessel repaired. The other is the record of that work: shift reports, timecards, inspection sheets, batch and lot records, change orders, and the paperwork a customer, auditor or regulator reviews later.
Your crews produce the first. Supervisors and office staff assemble the second, often at the end of the shift or after the job closes, by retyping what already exists somewhere else.
Most operations have already gone digital in places: timekeeping, tablets on the floor, an ERP, a shared drive, a WiFi network across the site. Yet the same numbers still get typed in two or three times: on paper at the machine, again into the timeclock, again into the ERP. Entered once, those numbers can fill in the shift report, the payroll check and the job record on their own.
Every sector runs on its own paperwork. These are the records the work runs on in each. The diagnostic shows which ones cost you the most hours.
Batch and device records, deviations and nonconformances, investigations, corrective and preventive actions, change controls, process validation and the release package. My experience is in pharmaceutical manufacturing, where I managed contract manufacturers at Takeda.
Travelers and work orders, inspection sheets and first article inspections, nonconformance reports, scrap and downtime logs, and shift reports.
Work orders and spec items, condition reports, change requests, timecards by job code and trade classification, Coast Guard, ABS and customer inspections, and the closeout package.
A Declaration of Inspection for every transfer, tank gauging and temperature readings, loss and gain reconciliations, and bills of lading for vessel, barge, rail and truck.
Bills of lading, receiving and OS&D (overage, shortage and damage) reports, cycle counts, labor by shift and the handoff between shifts.
Services
Each is a separate service. Some operations need one, some need all three. The intro call and the diagnostic show which fit yours. The order is always the same: fix the process first, then build a tool where the work still needs one.
01
Every engagement starts with research: interviews with the people who do the work, a review of your documents and data, and time on site watching the work run. The result shows where hours go and which fixes are worth making first.
What this includes
I trace jobs, orders and shifts from start to finish, time each step, map every handoff and set a baseline for each measure, using Lean Six Sigma methods including Kaizen, 5S and Kanban. Interviews and records fill in the tribal knowledge a single day on site cannot show.
Many of the biggest gains come from rules and routines: one controlled copy of every drawing, material staged before the shift, a clear owner for each handoff. I design these with the people who do the work and measure whether they hold.
What we measure
This is the right engagement if
A different engagement fits better if
02
Where software pays, I build tools that run on the data you already collect and cut manual data entry, and I help you choose new software when you need it.
What this includes
At the end of a shift, a supervisor records a two-minute voice note and takes a few photos. Those combine with timecard and job data to produce a draft shift report, inspection report or nonconformance report. The supervisor reads it, corrects it and signs it.
Reports, photos, hours and approvals for each job, order or lot collect in one place while the work happens. Hours get checked against job codes and pay classifications before payroll goes out. Open items such as change requests, deviations and customer approvals sit on one list with their status, and anything waiting too long gets flagged. At closeout, the package for the customer, auditor or regulator is already assembled.
A machinist, picker or welder pulls up the work instruction, drawing, procedure or equipment history on a phone at the machine, the rack or the hull. A material request goes from the floor to purchasing or the warehouse with the job already attached, and the status comes back to the same phone. Each morning, supervisors see what every job or order is waiting on: an approval, a part, a subcontractor, an inspection.
Put the hours and materials each job was quoted at next to what it actually took. Over a year of work, patterns show up: which kinds of jobs run over, which finish early, which machines or assets keep coming back for the same repair. The next quote starts from your own record as well as your estimator's experience.
Every vendor is adding AI to its pitch. I write requirements from how your work actually runs, compare the options against them, and test the leading choice on a real job before you commit.
Ground rules for every tool
What we measure
This is the right engagement if
A different engagement fits better if
03
New tools and processes hold only when your people know how to use them and the work around them is planned.
What this includes
Hands-on training for supervisors, office staff and crews on the tools they use and on AI in their daily work, so adoption holds after I leave.
I map which tasks change as new tools and processes arrive, which skills your people need next, and how to capture what your most experienced people know before they retire. Knowledge capture turns that know-how into procedures and records the next person can use. The plan covers tasks, skills and knowledge; hiring and pay decisions stay with you.
What we measure
This is the right engagement if
A different engagement fits better if
We talk through what you run and where time gets lost. If it makes sense, I then spend half a day on site at no cost and send you a written summary of what I find.
I interview the people who do the work, review your documents and data, and spend time on site. You get a ranked list of fixes, from process changes and training to new tools. The next section lists what it covers and what you keep.
I put the top fix in place for one crew, line or dock. Before it starts, we agree on the numbers it has to move, such as hours per shift spent on reports or days to closeout, and I measure them before and after. The diagnostic fee is credited toward the pilot.
After a working pilot, I keep tools running, train new staff and take on one new improvement each quarter.
Two to three weeks per site, through interviews, your documents and data, and time on site. I study the five areas together, because each feeds the others: a report gets retyped because two systems sit apart, and a job waits because its record is still open.
Who writes each shift report, timecard, inspection sheet and job record, how long each takes, and how often the same numbers get typed twice.
Where jobs, orders and lots wait on an approval, a part, a drawing or an answer, and for how long.
What your ERP, timekeeping, shared drives and floor devices already capture, and where the data stops connecting.
How long it takes a crew member to find the current drawing, procedure or equipment history.
Which tasks depend on one or two experienced people, and what they know that lives only in their heads.
A ranked list of fixes. Each one shows the hours it saves, what it costs, the effort it takes from your people, and whether it needs software at all. The list and the baseline measures are yours to keep, whatever you decide next.
One crew, one line, one kind of report. It spreads across the operation after it has earned its place.
Supervisors, office staff and crews test early versions on real jobs, and their corrections shape the final version. A change is finished when the people doing the work use it without help.
Each tool and process comes with plain-language instructions and training for the people who run it, so your team understands it and can keep it running.
I'm based in Bayonne, New Jersey, and work on site with operations across New York City, New Jersey and Eastern Pennsylvania.
I come from operations. I started my career in pharmaceutical manufacturing and supply chain management at Takeda, a Japanese drug maker founded in 1781. I managed contract manufacturers for drug product and for packaging and labeling, and worked with distribution centers and contract QC testing labs. I wrote batch records, ran manufacturing investigations and change controls, and worked on two drug launches. In that work, every batch, every hour and every change is documented before product ships.
At Takeda I earned a Lean Six Sigma Green Belt. The training drew on the Japanese methods behind lean manufacturing, including Kaizen, 5S and Kanban. Every process improvement project I worked on started the same way: mapping how the work actually moved before changing anything.
More recently I built 48 AI tools for a software company, Thought Industries, and trained its teams to use them. Each one connects to systems the company already runs and takes on work people had been doing by hand.
My work is new to many of the industries on this page. Documented, regulated production is work I know well, and I learn each operation by talking with the people who run it, reading its records and watching the work.
Often you need less than vendors suggest. Many fixes are process changes, and most tools run on the systems you already have, such as your ERP, timekeeping, email and shared drives. When new software is the right answer, I help you choose it.
No. I'm a consultant. When new software is the right answer, I write the requirements, compare the options and test the leading choice on a real job, and you buy it directly from the vendor. Every tool I build belongs to you.
NJMEP and DVIRC are the Manufacturing Extension Partnership centers for New Jersey and Southeastern Pennsylvania, part of a national network backed by NIST. Through their staff and vetted partners, they cover lean, quality systems and ISO certification, safety, cybersecurity compliance and workforce training. My work sits alongside theirs: AI tools that run on the data you already collect, the reports and records around the work, and planning for how tasks change as new tools arrive. For ISO certification, OSHA training or CMMC readiness, they are the right first call.
You do: the tool, its instructions and its data. Your team can run it without me.
We agree during the diagnostic which data each tool reads, where it is stored, and which AI services it uses. Tools run in accounts your company controls.
I know regulated production and documented work from pharmaceutical manufacturing and supply chain management at Takeda. I also know how products get built and adopted. I've worked on product development for software, including AI tools for two software companies, and for consumer products at Pilot44 with PepsiCo, Kimberly-Clark, Nestlé and Whirlpool. That work trained me to scope a tool, test early versions with the people who will use it and keep improving it until they rely on it. I learn the specifics of your operation the same way I start every project: by talking with your people, reading your records and spending time where the work happens.
It reads the reports, photos, schedules and records you already produce, drafts from them and checks them. A person reviews and approves the result. Part of my work is deciding where AI helps and where a simpler fix does the job.
Tell me what you run and where time gets lost. We'll set up a 30-minute call and decide together whether a visit makes sense.
Request a 30-minute callbrendan@thinkarc.io