This is our first blog post, so it seems fair to use it to explain what we're building, why, and what you can expect from us going forward, rather than jumping straight into the usual content.

The short version: QuantCAD is an AI-powered 3D CAD platform for people who need to go from idea to 3D model without a two-week onboarding process and a four-figure software licence. The longer version requires some context.

The problem we're actually trying to solve

Hardware product development has a step that's quietly expensive in a way most people don't talk about directly. The step between "I have an idea" and "here's a 3D model other people can react to."

For a long time, that step required either deep CAD expertise or access to someone who had it. You needed to know the software, understand the conventions, manage the file formats. If you didn't have that, you either hired someone who did, relied on whoever on the team happened to know the tools, or worked around the problem entirely, which usually meant staying in 2D sketches and descriptions longer than was good for the project.

None of this is dramatic. It's just a quiet tax on every early-stage hardware project. And like most quiet taxes, it gets absorbed as a cost of doing business without anyone seriously asking whether it needs to exist.

We think it doesn't need to exist. Or at least, not at this scale.

What QuantCAD actually does

QuantCAD is a 3D CAD platform that runs in the browser. You describe what you want to build in plain language, with whatever level of detail you have, and QuantCAD generates a working 3D CAD model, built as B-rep solids, as a starting point.

The model isn't a static render. It's real geometry you can inspect from any angle, refine with further prompts or direct edits, and export in the formats that downstream workflows actually need: STEP for manufacturing, STL for printing.

A few things that matter to us about how it works:

  • No installation. QuantCAD runs in a browser tab. No workstation setup, no licence keys, no compatibility issues between operating systems. The barrier to getting started is opening a link.
  • The output is editable, not final. AI-generated geometry is a starting point, not a deliverable. We've built the platform to make iteration fast, so you can refine, adjust, and correct rather than regenerating from scratch every time something needs to change.
  • Built for teams from the start. Version control and lifecycle state management are part of the platform, and team collaboration is available on Enterprise plans. Most hardware development is a team exercise, and the platform is designed around that reality.
  • Export-ready output. The geometry you generate in QuantCAD can go directly into manufacturing quotes, simulation tools, or further detailing workflows. What you export is a CAD model you can keep working on.

Who we're building this for

Honestly? A fairly wide range of people, because the underlying problem, the expensive gap between idea and 3D artifact, affects a lot of different roles.

Product designers who think in 3D but aren't CAD experts. Founders who need to show something physical to investors or manufacturers. Engineers who want to move faster at the concept stage without shortchanging the rigorous part of their process. Small teams who can't justify a CAD seat for every person who needs to look at or react to a design.

And experienced CAD engineers who just want to stop spending the first two hours of a new project staring at an empty viewport.

What you can expect from this blog

We'll use this space for things that are actually useful, not for content marketing dressed up as engineering advice.

That means workflow posts about how to get better results from AI-assisted CAD. Honest write-ups when we ship something new: what it does, why we built it, and what its limitations are. Some industry thinking about where this technology is going and what's worth paying attention to. And case studies when we have them: real designs built with QuantCAD, real process, real output.

We're not going to publish posts designed to rank for keywords. We're going to publish things we'd actually want to read ourselves.

One more thing

We have a clear point of view on AI in engineering: it shouldn't replace engineers, it should free them up. The parts of the design process that consume time without requiring expertise, such as generating early-stage geometry, producing documentation drafts and exploring parameter variations, are the right targets for AI assistance. The parts that require judgment, accountability, and contextual knowledge should stay with people.

That's the philosophy behind QuantCAD. We think it's the right one, and we're building the platform to reflect it.

If you're here because you're curious about what we're building, welcome. If you have a design challenge you want to test against the platform, go ahead and try it. And if you have feedback about the product, or about what you'd find useful in this space, we genuinely want to hear it.