Back to BlogTLDR: Onboarding a CAD designer is expensive because it is paid in senior technician time, which is the same capacity the lab is short of. Structure it around reviewing real cases rather than building from scratch, keep a written standard for what good looks like, and stage the work by case difficulty rather than by calendar.
Why is training so expensive in a lab specifically?
Because the teacher and the bottleneck are the same person. Most dental laboratory technicians learn on the job, and the only people qualified to teach digital denture design are the senior designers already carrying the production load. Every training hour is an hour of output you gave up.
This is not a soft cost. If a senior designer spends a quarter of their week teaching for three months, the lab has spent roughly three weeks of senior production capacity, at exactly the moment it was short. Labs routinely make this decision without costing it, then wonder why the backlog grew during onboarding.
It is also why the formal route is thin. The American Dental Association lists 14 accredited dental laboratory technology programs nationally, and typical entry-level education for the occupation is a high school diploma. The trade is mostly transmitted person to person, and that channel has a hard capacity limit.
Start with review, not with building from scratch
A trainee who reviews and corrects real designs sees far more cases per week than one building each from nothing, and case variety is what builds judgement. Reverse the usual order: teach them to recognise a good design and a bad one first, then teach them to produce one.
The conventional sequence has a trainee construct cases slowly from the beginning while a senior corrects the output. It works, and it is slow, because a trainee might complete three or four cases in a day and therefore encounter three or four situations.
Flip it. Put proposed designs in front of the trainee and have them say what they would change and why, with a senior checking the reasoning rather than the geometry. A trainee can work through many more cases this way in the same time, and the thing being trained is judgement, which is what actually takes years to build.
This maps directly onto how AI-assisted workflows already run. In SmileShape, SmartCAD proposes a full-arch design and a qualified technician reviews, adjusts and approves it. A trainee working alongside a senior on that review step is doing the real job from week one, at supervised volume, rather than simulating it.
Write down what good looks like
Most labs carry their design standard in the heads of two or three people, which makes every review a negotiation and every disagreement personal. A written standard turns training into checking work against a document rather than against a colleague's memory.
A usable standard is short and specific. Cover at least:
- Trim line criteria, and the cases where your lab deviates from the default.
- Occlusal plane conventions and how you handle the awkward ones.
- Setup preferences your labs' referring offices expect, by account where they differ.
- Thickness and strength minimums for the materials you actually run.
- What gets escalated rather than decided: the list of situations where a trainee should stop and ask.
- What gets a call to the referring office rather than a judgement call in house.
Writing this down takes a senior designer perhaps two days and it is the highest-return two days in the whole onboarding programme, because it is spent once and used by everyone afterwards.
Stage by difficulty, not by week
Calendar-based onboarding advances people on a date. Difficulty-based onboarding advances them when they have demonstrated they can handle a class of case. The second produces fewer remakes and gives you a real answer when someone asks whether the new designer is ready.
- Stage 1. Review only, on straightforward complete denture cases. Trainee states what they would change; senior decides. Exit when their calls match the senior's consistently.
- Stage 2. Review and adjust on the same case class, with a senior approving. Exit on a clean run of cases with no corrections that a standard would have caught.
- Stage 3. Full responsibility on that case class. Second review on a sample rather than on everything.
- Stage 4. Repeat stages 1 to 3 for the next class: partials, immediates, implant-supported full arch.
What to measure
- Cases reviewed per week during the review stages. This is the exposure that builds judgement, so it is the leading indicator.
- Correction rate, and more usefully the trend. Flat is a signal to change how you are teaching, not to wait longer.
- Escalations. A trainee who never escalates is guessing; one who escalates everything has not been given a standard clear enough to act on.
- Senior hours spent. Track it honestly, because it is the real cost of the programme and the number that tells you whether the structure is working.
Related reading
- the dental CAD designer shortage in numbers
- what human review looks like in an AI-assisted denture workflow
- hire, outsource, or add AI-assisted design
Frequently asked questions
How long does it take to train a denture CAD designer?
It depends far more on case exposure than on calendar time, which is why advancing people by demonstrated competence on a case class works better than advancing them by week. Review-first onboarding increases exposure per week substantially.
Why start a trainee on review instead of building designs?
Because judgement is the slow thing to build and review exposes a trainee to many more cases per week than construction does. Once they can reliably tell a good design from a poor one, producing one is a much shorter lesson.
What should be in a lab design standard?
Trim line criteria, occlusal plane conventions, per-account setup preferences, material thickness minimums, the list of situations that get escalated, and the list that gets a call to the referring office.
Does AI-assisted design make training easier or harder?
It changes what is trained. Because the technician's job is reviewing, adjusting and approving a proposal, a trainee working that step alongside a senior is practising the real task at supervised volume from the start.