Limitless Labs, a Physical AI Platform for Mechanical Manufacturing, Announces $20 Million Series A Funding
Limitless Labs, a physical AI platform targeting the mechanical manufacturing sector, announced the completion of a $20 million Series A funding round, co-led by Dell Technologies Capital and Square Peg, with participation from Grove Ventures, Meron Capital, and Kinetica. The funds will be used to accelerate product deployment and international market expansion.
The company claims to be the world's first "intelligent agent-based physical AI platform" for CAD/CAM scenarios in mechanical manufacturing. The AI agent is directly embedded in the CAD/CAM software used by engineers, helping manufacturing companies capture, standardize, and scale the process expertise of experienced CNC programmers, potentially reducing CNC programming time by up to 50%.
Since emerging from stealth mode, Limitless Labs has expanded from pilot projects to full production deployment, with clients including Blue Origin, Cadillac F1, Sandvik, and Iscar. High-end manufacturing companies in aerospace, defense, racing, and industrial machinery are beginning to transition some of their process knowledge from a "master-apprentice" model to "AI agent reuse." Funding and computing power are concentrating on CNC and precision manufacturing lines with high margins and complex processes.
Source: Public Information
ABAB AI Insight
From historical behavior, Limitless Labs initially entered the CNC niche process scene as LimitlessCNC, essentially penetrating the "most difficult and scarce" CNC programming segment first, before expanding to a broader CAD/CAM physical AI platform. This path differs from traditional general-purpose AI PLM or industrial software, prioritizing the elimination of the highly monopolized "black box experience layer" held by a few senior process engineers, similar to how EDA tools gradually replaced some work of senior layout engineers with automated layout and routing.
In terms of capital pathways, the joint investment by Dell Technologies Capital, Square Peg, Grove Ventures, etc., indicates that this sector is viewed as a mid-to-long-term infrastructure bet akin to an "industrial version of Copilot + EDA": investors are betting that by distilling the high-salary CNC/process engineer experience into replicable AI agents, the marginal efficiency of each production line can compound continuously, with the potential to extract value in every design iteration and processing task, rather than just selling a software license.
In industry comparison, Limitless Labs' current positioning is closer to a "code compiler + optimizer for the physical world": CAD is the source code, and CNC programs are the target code. The traditional manual processes of path planning, tool selection, and processing strategy optimization are gradually being taken over by AI agents, which is very similar to the evolution of software development from hand-written Makefiles to IDE auto-completion, and then to CI/CD pipelines; only this time it occurs on machine tools and cutting tools, where the cost of errors is wasted materials and machine downtime, rather than simple online bugs.
Structurally, this marks the starting point of the shift in industrial manufacturing from human experience pricing power to AI toolchains: as key clients like Blue Origin, Cadillac F1, Sandvik, and Iscar gradually hand over core processes to embedded AI agents, the delivery cycles, yield rates, and costs of high-end CNC projects will increasingly be determined by platform algorithms and data network effects, rather than the individual capabilities of a single process engineer; in the long run, the role of top engineers will shift from "programming personally" to "training and supervising AI process agents," while the physical AI toolchain itself will be standardized and globally disseminated.
ABAB News · Cognitive Law
When experience is written into models, masters do not disappear, but daily wages no longer determine productivity.
The world of code has compilers; the world of factories will eventually have process compilers too.
Engineers sell labor hours as physical work, while what they sell to AI is scalable process structures.