Four years after its last big raise, the CNC automation startup is chasing scale in a labor-short U.S. market.
Ask Theo Saville why CNC programming resists automation and he reaches for a number: there are more ways to cut a typical part than there are atoms in the universe. That combinatorial mess is what CloudNC has spent a decade attacking. On Wednesday, the U.K. company added a $20 million Series B extension, pushing its lifetime total to $128 million.
CNC machining turns raw material into precise parts for cars, defense systems, and consumer hardware. Before anything gets cut, a programmer decides how the part should be held, which tools to use, and at what speeds. Conventional CAM packages make that work faster, but they still wait for a human to specify the strategy.
CAM Assist plugs into those packages, including Autodesk Fusion and Mastercam, and drafts the first pass itself. It picks tools, approach directions, and cutting feeds and speeds, then writes the code that tells the machine what to do. The programmer reviews, edits, and approves. Saville frames it as an expert sitting beside the programmer rather than a replacement for one. The repetitive setup thinking gets automated. The judgment stays human.
More than 1,000 machine shops use CAM Assist. Eighty percent of the customer base is in the U.S., and CloudNC runs the whole operation with 80 employees. Scaling adoption is now the entire job, which is why Saville says this was the moment to raise again. His last major round closed four years ago.
Nimble Ventures led the extension. Calculus Venture Capital, Entrepreneurs First, and LM Capital, Lockheed Martin’s venture arm, joined. The money goes toward go-to-market operations, existing and new markets, and product development.
Quote Agent launches next month. It sizes up the cost and risk of a new job so a shop can accept or turn it down quickly. Saville ties the need to two forces pulling against each other: American manufacturers are reshoring work while skilled machinists stay scarce. Shops have to quote faster and program faster using the people and machines already on the floor.
If you sell into manufacturing, the buying trigger is moving from capacity to output per head. Shops are not hiring their way out of the labor gap, so what lands is hours recovered per programmer, not feature lists. Note the cap table too: Lockheed Martin’s venture arm is in, which signals that defense supply chains now treat CNC bottlenecks as strategic.
The harder question is what a shop does with the hours it gets back.
