And the honest answers I keep coming back to.
In hundreds of conversations with millwork and cabinet shop owners over the last few years — different sizes, different software, different markets — the same handful of questions come up again and again.
Here are the five I hear most, with the honest answers I give. And watch for the pattern, because it’s the most useful thing in this article: almost none of these questions are really about what they sound like.
1. “How do I get more capacity without hiring?”
If you’re drowning in work right now, hiring someone is not going to fix your problem anytime soon.
Onboarding a new employee takes three to six months, minimum, before they’re actually productive. And getting them there costs you the one thing you don’t have when you’re swamped: your current team’s time to train and onboard them. So the very moment you most need relief, hiring adds load before it ever removes any.
That’s why more capacity, fast, usually has to come from somewhere other than a new W-2 — whether that’s better process, offloading the lower-value work, or bringing in an experienced team that can jump in without needing months of ramp. The goal isn’t just more hands. It’s more hands that don’t cost you a quarter of a year and half your team’s attention to become useful.
2. “Where is the next generation of drafters and engineers going to come from?”
This is the quiet crisis in our industry, and almost everyone feels it.
The people who hold the deep knowledge are retiring. The generation that used to train juniors is aging out right alongside them. And meanwhile every shop is chasing the same shrinking pool of experienced, day-one-ready drafters — who mostly don’t exist anymore. If your plan is to recruit your way out of this, you’re fighting over a resource that’s disappearing.
The shops that will still be here in ten years are the ones that stop recycling talent and start building it. There are two pipelines, and you want both: upskill people already on your shop floor who know how you build — they’re future drafters who just haven’t learned CAD yet — and bring in CAD-trained grads or interns who can learn millwork from you. They meet in the middle.
It doesn’t take a formal program to start. Pick one promising person, give them Friday afternoons to learn the software, a six-month window, a clear goal, and a small bump in pay when they hit it. That’s not a training cost — it’s insurance on your pipeline. And it sticks, because people stay where they’re invested in. Career development, not pay alone, is the number-one reason good people leave.
3. “Is AI going to replace drafting?”
I think the question is framed wrong. The real question isn’t whether AI replaces drafters — it’s what happens when the drawing part gets easier.
Right now, the tools aren’t there. They can spit out generic 2D or 3D, but they don’t understand casework logic, and they don’t integrate with the CAD/CAM software our shops actually run on, like Microvellum or Cabinet Vision. That will change.
When it does, AI becomes a translator — a first-draft generator that turns plain-English fabrication intent into real work. But here’s the catch, and we’re already watching it play out with software engineers using AI on code: you can get something like 10x the output, but only if you understand what you’re looking at well enough to validate it. AI-generated work carries a higher defect rate, and someone still has to know the difference between technically correct and actually buildable.
So the drawing part becomes table stakes. The thinking part — the fabrication knowledge, the standards, the judgment — becomes the job. That knowledge doesn’t get automated. It gets amplified. Which means the shops positioned to win with AI are the ones that already have their standards and process figured out. Don’t wait for AI to rescue a broken process. It won’t.
4. “Why am I doing the work twice?”
A lot of shops are stuck in a loop: draft it in 2D for submittal, get it approved, then re-engineer the whole thing in their CAM software for production.
The true cost here isn’t mainly the extra engineering labor — although that’s real. It’s the schedule delays that come from not being able to move quickly from approved drawings into engineering and production. Every day lost in that handoff is a day added to your lead time.
One nuance worth understanding: it’s perfectly valid to design your workflow this way on purpose — submittals in 2D, then an engineering stage that puts the products into your CAM software. That’s a legitimate choice. But if your normal workflow is to draw products in your CAM software from the start, then everyone touching that work — in-house or outsourced — should be doing the exact same thing. The mistake is letting a double-handling step creep into a process that never needed one. Match the workflow you actually run.
5. “Should I switch software?”
The vast majority of the time I get this question, it’s coming from an owner who’s frustrated with how long it takes their engineering team to get through work. And virtually every time, changing software won’t solve that problem.
Each of the major options on the market — Microvellum, Cabinet Vision, Mozaik — can be effective. If your company has already spent significant time both configuring and learning a platform, the cost of switching usually isn’t worth the squeeze. You’d be trading a known system for a fresh learning curve and re-doing all that configuration — and your real bottleneck would follow you right over.
So I almost always recommend the same thing: before you evaluate software, evaluate your processes, your team’s current training and knowledge level, how they could be using the software better, and the workflows between engineering and the rest of your organization. That’s where the speed you’re looking for actually lives.
The pattern
Capacity. Talent. AI. Rework. Software. On the surface, five different problems.
Underneath, they mostly come back to the same place. Hiring won’t fix a process that can’t onboard people. Recruiting won’t fix a talent pipeline you never built. AI won’t rescue standards that live in one person’s head. A new platform won’t fix a workflow that doubles its own work. The shops that handle all five well have one thing in common — they run on systems, not on heroics.
So before you hire, chase talent, wait on AI, or switch tools, ask the harder question: is the real constraint capacity and software, or the way work and knowledge move through my shop?
Most of the time, it’s the second one. And that’s good news — because it’s the one you can actually build.
Jacob Edmond
CEO