A generic AI product design generator gives you a hundred attractive renders and no way to tell which of them can be built. The useful version knows your brand language, your envelope and your process, and says so on every concept.
All Product & Industrial Design →
Generative concepting means using image and 3D models to produce many early product concepts from a brief, then sorting them against the constraints that will matter later, so a design team explores widely without wasting a week on forms that cannot be made.
The image tools are already good. Midjourney, Stable Diffusion and Vizcom will turn a sketch into a convincing render in seconds. What they do not do is remember that the product must fit a 90 mm envelope, ship in two injection-moulded halves and look like the rest of your range. Those constraints live in the brief and in the heads of the senior designers, and the concepts drift away from them within an afternoon.
So the work is mostly around the generator rather than inside it: a reference set drawn from your own archive and brand book, a brief written as explicit constraints, and a sorting step that reads every concept against them and says which ones survive and why.
Before anything is built, we take a past project where you know which concepts made it through and which were dropped, and run the same brief. The sorting step is judged on whether it would have flagged the concepts your team rejected for manufacturing or envelope reasons. That is a measurable question; whether a render is beautiful is not, and we leave it to the designers.
Two practical limits are worth stating. Rights in AI-generated imagery are unsettled in several jurisdictions, so we keep the generated concepts as inputs to human design work rather than as deliverables in their own right. And a concept that resembles an existing product should go through design search before anyone grows attached to it.
written as explicit constraints
your archive and brand language
sketch and text to render
envelope · parts · finish
and models in CAD
breaks a hard constraint
For early form and mood, image tools such as Midjourney, Stable Diffusion and Vizcom are strong. For engineered parts, generative design in CAD tools such as Fusion or Creo is a different and complementary method. The gap between them is the constraint check: knowing which attractive concept fits your envelope, process and brand. That is the part we build, around whichever generator suits your team.
Not on its own. Image models produce pictures and rough meshes, with no knowledge of draft angles, wall thickness or tolerances. A generated concept is a starting point for a designer modelling in CAD. Manufacturability is checked later, on the real model, by a rule-based design-for-manufacture step and an engineer.
It depends on the jurisdiction and the tool terms, and the law is still moving. Copyright offices in several countries have refused protection for purely machine-generated images. That is why we treat generated concepts as inputs to human design work, keep a record of who developed each direction, and recommend your IP counsel confirm the position before you rely on a generated image commercially.
What we built, what stage it is at, and a link to the live thing wherever there is one.
yourOS is an operating system with an AI at its heart. You talk to it like a person, out loud or typed. It makes your calls, plans your days and works on your projects with a crew of agents, on your own private server. Our own product, built from the first commit; the waitlist is open.
Website, backend and app assets for Sonora, a free AI sound-therapy app on iOS and Android published by Cognitive AI.
Brand, website, prototype, apps on five platforms, the AI backend and the algorithm-intelligence pipeline for Podswap, a social growth engine that runs without a human in the loop.
In closed beta: Roxy, a private AI assistant with permanent memory running on infrastructure the user owns. Brand, site, product, backend and conversational design.
Live at coherence.com: brand identity generated with AI agents we directed, the motion system and website for the group behind Sonora, AboutAI, WakeUp and MudShadows. Agent work remains in build.
Live at domainbooth.com for DomainBooth, a premium domain brokerage: a new animated mark, every page redesigned and rebuilt with designed mobile layouts, and a custom CMS for its sales record, featured names, press and enquiries.
Bring a past brief and the concepts that survived it; we will show what the sorting step would have said.
30-minute technical call · no deck