We know the constraints before you explain them.
Every industry below has shipped product behind it. Each one has its own regulatory envelope, its own cost drivers, and its own way of failing. Starting from the one closest to your problem saves months.
Robotics & Automation
Automation hardware fails on the boring parts: a fixture that deflects under load, an enclosure that traps heat, a cable route nobody modeled. The control software is usually the part that gets attention and the mechanical detail that causes the downtime.
Where this runs: XCAD is deployed in automated parking garages and warehousing logistics, alongside machinery manufacturing. See the program →
Consumer Hardware & IoT
Consumer hardware has the narrowest margin for error and the widest audience for it. Cosmetic criteria have to be agreed before tooling, not after the first shot, and the app is part of the product whether the roadmap says so or not.
Typical entry point: a working prototype with no path to a cosmetic standard. How we handle tooling →
Medical Devices
In medical work the documentation is part of the deliverable. A fixture that works but can't be traced is a fixture that fails an audit. So inspection criteria and records are scoped at the same time as the part.
Shipped: an orthopedic institution sent CT data and had a printed bone model in one day, at 0.05 mm. See the program →
Automotive & Mobility
Automotive work rewards process discipline over cleverness. Cycle time, repeatability and tolerance stack decide whether a part is viable long before aesthetics enter the conversation.
Typical entry point: a bridge-tooling decision where the volume assumption is still moving.
Industrial Equipment
Industrial equipment is where design and plant reality collide. The machine has to fit a building that already exists, be serviceable by the people already employed, and hold up for a decade of shifts.
Typical entry point: a new line where the layout is being decided on a spreadsheet. How we model it →
Energy & Cleantech
Energy manufacturing is capital-intensive and unforgiving of rework. The highest-value intervention is usually before construction — modeling the plant so layout, routing and interference are settled while changes are still cheap.
Shipped: a four-metre governance model for an environmental engineering firm, and a digital factory deployment for a solar manufacturer. See the program →
Media & Entertainment
Live media is a latency and reliability problem wearing a creative costume. The studio, the encoding pipeline and the delivery layer have to be designed together, because the failure always shows up at the seam between them.
Shipped: live streaming, virtual studio and audience tipping running on one stack, not four vendor integrations. See the program →
Software & Platforms
Independent software vendors carry a compatibility tail that constrains every architectural decision. The work is rarely greenfield. It's moving a trusted product forward without breaking the files, integrations or workflows its customers depend on.
Shipped: a design toolchain from drafting through CAE optimization to CAM toolpath — a part taken from 5.7 to 3.0. See the program →
Not on the list?
A track record, not a limit.
We have shipped in industries that don't have a clean category name — cultural products, scale models, precision instruments, and a few things we can't describe in a public page.
What transfers between industries is the method: lock intent, review manufacturability, prove it physically, then ramp with a control plan. The vocabulary changes. The sequence doesn't.
Tell us what you are building.
Send a drawing, a repo, or a paragraph. You will hear back from an engineer, not a sales development rep, within one business day.