TLDR: SmileShape reports that a skilled CAD designer can move up to 30 cases through initial AI-assisted design in an hour by running several at once. That figure counts first-pass design proposals entering review, not finished dentures leaving the building. Every case still stops at a qualified technician who reviews, adjusts and approves it.

What does 30 cases an hour actually count?

It counts cases that reach a reviewable first-pass design inside one designer-hour, with cases running in parallel rather than one after another. It does not count finished, approved, production-ready cases. Those still move at the speed of human review, which is the point.

Throughput claims in this industry are slippery because nobody agrees on the unit. A number that sounds like finished dentures and turns out to mean something upstream is the fastest way to lose a lab owner's trust, so it is worth being precise.

The unit here is the initial design proposal. SmileShape's SmartCAD takes a validated scan and proposes a full-arch design, including margins and anatomy, in roughly two minutes. A technician then reviews that proposal, adjusts it, and approves it. The 30 figure describes how many proposals a single designer can have generated and queued for their own review inside an hour. The review is separate, and it is human.

Why running cases in parallel matters more than per-case speed

A two-minute design step run sequentially gives you 30 cases an hour only if the designer does nothing else for those two minutes. Run in parallel, the designer starts the next case while earlier ones are still processing, so the hour is spent on judgement work rather than waiting.

Work out the arithmetic and it stops being a marketing number. Thirty cases at roughly two minutes each is sixty minutes of processing. If a designer had to sit and watch each one finish, the ceiling would be exactly 30 and the designer would have contributed nothing but patience.

Parallel execution changes what the hour is for. The designer submits, moves on, and comes back to a queue of proposals ready for review. Their hour goes into the parts of the job that need a trained person: reading the prescription against the scan, judging the trim line, confirming the occlusal plane, deciding whether an anatomy proposal suits this patient's case.

This is also why the number is capacity rather than speed. Nothing about one case got dramatically faster for the technician. What changed is how many cases one technician can carry at once.

What does the technician still do?

Three of the five workflow stages require a trained technician to review and approve before the case advances. The technician approves the trim line, confirms the occlusal plane, and reviews, adjusts and approves the final design before it goes to production. No case skips those gates.

The full path a case takes through the platform:

1. The clinic scans and submits the case, with no change to the clinician's existing workflow.
2. SmartRX reads the prescription on upload and populates the case record, cross-referencing it against the scan files.
3. SmartScan checks the scan and flags quality issues for the technician: distortions, missing mesh, contamination artifacts, unclear margins.
4. The technician reviews and approves the trim line, then confirms the occlusal plane.
5. SmartCAD proposes the full-arch design.
6. The technician reviews, adjusts and approves the design, then sends it to production.

Steps four and six are hard stops. The software does not advance a case past a technician who has not approved it, and the design that reaches manufacturing is the one a qualified person signed off on. SmileShape is not a diagnostic or treatment-planning tool and does not make clinical judgements.

What the number does not mean

• It is not 30 finished dentures an hour. Review time is real and it belongs to a person. A lab that plans capacity as though the design proposal were the finished case will overcommit.
• It is not a replacement for CAD designers. The figure is stated for a skilled designer. Remove the designer and the number has no meaning, because there is nobody to approve anything.
• It is not a claim about how good the proposals are. Throughput is a business measurement. Whether a given proposal suits a given case is a judgement the technician makes, every time.
• It is an upper figure, not an average. "Up to" is doing real work in that sentence. Case mix, scan quality and how cases arrive through the day all move the practical number.

How do you tell whether your lab can use capacity like this?

Capacity only helps if you have demand waiting for it. The labs that see the most from parallel design are the ones already turning work away, sitting on a backlog, or paying per arch for outsourced design. If your designers have idle afternoons, faster design solves a problem you do not have.

Three questions worth answering honestly before you evaluate any design tool on throughput grounds:

• Is design actually your constraint? Measure where cases sit longest. In plenty of labs the wait is on incoming information, not on design, and adding design capacity in front of an intake bottleneck just moves the queue.
• Do cases arrive in bursts? Parallel processing pays most when Monday brings a week's work at once. A steady trickle benefits less, because there was never a pile to work through.
• What do you currently pay per arch? Labs paying an outside service per arch have a clean number to compare against. SmileShape reports labs reducing CAD design cost by up to 60 percent against outsourced rates of up to about $40 an arch.

SmileShape is built for mid-to-large labs running roughly 1,000 or more cases a month, which is the volume at which parallel design capacity starts to matter more than per-case speed.

Related reading

• what human review looks like in an AI-assisted denture workflow
• the digital denture workflow end to end
• work out your real cost per denture design