Limitless Labs Raises $20M for Physical AI

Limitless Labs Raises $20M to Bring Physical AI Into the Heart of Precision Manufacturing: Already Powering Blue Origin Rockets and Cadillac's F1 Car

There is a version of artificial intelligence that most people never think about. Not the kind that writes emails, summarises documents, or generates images, but the kind that figures out how to cut a titanium aerospace component to within a few microns of tolerance, or programs the machining sequence for a structural part that will one day sit inside a rocket hurtling toward the edge of space. This is the world that Limitless Labs was built for, and the market is starting to take serious notice. The Tel Aviv-based startup, founded just two years ago in 2024, has closed a $20 million Series A funding round, announcing on June 16, 2026 that the round was co-led by Dell Technologies Capital and Square Peg, with additional participation from Grove Ventures, Meron Capital, and Kinetica. The latest raise brings the company's total funding to $27.3 million, and comes at a moment when Limitless Labs is not pitching pilots or running proof-of-concept trials — it is already in full production with some of the most demanding industrial organisations on the planet, including Blue Origin, Cadillac Formula One, Sandvik, and Iscar.

For a company that only emerged from stealth recently, the production-stage relationships with clients like these are remarkable by any standard. Blue Origin builds rockets. Cadillac's Formula One programme manufactures components that must perform flawlessly at speeds exceeding 300 kilometres per hour. Sandvik is one of the world's most respected names in cutting tools and machining technology. Iscar, an Israeli-based subsidiary of Berkshire Hathaway's IMC Group, is a global leader in metal cutting tools. These are not organisations that experiment recklessly with unproven technology. The fact that all four have moved Limitless Labs from evaluation to live production deployment says more about the platform's maturity than any press release could. According to the company, its system has already helped cut CNC programming time by up to 50% across these production environments, a figure that, if sustained at scale, represents a genuinely transformative shift in manufacturing productivity.

Bridging the Widening Skills Gap in Precision Manufacturing

To understand why Limitless Labs exists, and why the problem it is solving has become urgent enough to attract blue-chip investors, it helps to spend a moment with the numbers. Close to a quarter of all manufacturing workers in the United States are currently 55 years of age or older. That statistic, taken alone, might not seem alarming. But paired with the reality that there are already around 409,000 unfilled manufacturing positions in the US today, and that this shortfall is projected to balloon to 1.9 million by 2033, it becomes much harder to ignore. Nearly all manufacturers, around 97% according to industry surveys, now identify knowledge retention as their single biggest operational concern. This is not just a worry about losing experienced workers to retirement. It is a deep structural anxiety about the disappearing institutional memory that keeps advanced manufacturing functional.

The reason this matters so much in precision machining comes down to something the industry calls tribal knowledge. When an engineer with 30 years of experience looks at a complex CAD model, they do not just see geometry. They see the accumulated logic of thousands of previous jobs: which cutting tools will deflect under certain loads, which sequence of operations minimises vibration, which approach to a particular feature type is likely to cause chatter or surface defects, which parameters will hold tolerance across a long production run. That knowledge lives in their heads, in the notes they have made over decades, in the conversations they have had with colleagues on the shop floor. It is rarely written down in any systematic way. When those engineers retire, that knowledge disappears with them, leaving younger machinists to rediscover lessons the hard way, often at the cost of scrapped parts, lost time, and expensive rework.

It was this specific problem that David Priev, Assaf Peleg, and Shahaf Finder set out to solve when they founded Limitless Labs in 2024. Rather than approaching manufacturing AI as a generic productivity tool, they built their platform with a singular focus: capturing the expertise of veteran machinists, standardising it, and scaling it across entire organisations so that the next generation of engineers could benefit from the accumulated knowledge of the best minds in the industry. As David Priev put it in announcing the funding round, the manufacturing world does not just need more automation — it needs a better way to capture and scale the expertise that still lives inside the heads of a relatively small number of experienced machinists. He went further, articulating the company's broader ambition: a belief that the next major AI platform will be built for the physical world, and that this transformation starts with giving manufacturers a way to scale their best knowledge across every new part and every new engineer.

The Physical AI Foundation Model: What Sets It Apart from the Crowd

The technology at the core of Limitless Labs is what the company calls its Physical AI Foundation Model. This terminology is worth unpacking because it signals something meaningfully different from the way most industrial AI tools have been developed. The dominant paradigm in AI over the past several years has been to train large foundation models on massive datasets of text, code, or images, and then adapt those models to specific tasks through fine-tuning or prompt engineering. The results in purely digital domains have been impressive. But when those same approaches are applied to the physical world of manufacturing, they tend to fall short in ways that are difficult to paper over.

Manufacturing is governed by physics. Metal has specific properties that dictate how it behaves under cutting forces. Tool geometry interacts with workpiece material in ways that create heat, deflection, and wear patterns that can be modelled mathematically. CNC machines have axes, spindle speeds, feed rates, and rigidity constraints that are defined by the physical realities of the equipment, not by anything that can be learned from reading text on the internet. Building a foundation model that actually understands these realities requires training on a fundamentally different kind of data: the physics of metal cutting, CAD geometry, machining processes, tooling requirements, and real machine constraints. This is what Limitless Labs has done, and it is what separates the platform from the wave of general-purpose AI tools that have been retrofitted for industrial applications with mixed results.

The platform's CAM Agent integrates directly into the computer-aided manufacturing software that engineers already use: Mastercam, Siemens NX CAM, and PTC Creo. This integration point is strategically important. Rather than asking engineers to adopt a new tool or change their workflow, the system works inside the environments they already know. When a CAD file is loaded, the platform identifies the geometric features of the part, recommends appropriate tooling, sequences the machining operations in an optimal order, generates toolpaths, and helps produce a production-ready CNC program. Throughout this process, the engineer remains in control of the final output. The AI is not replacing their judgement — it is augmenting it with the institutional knowledge of the organisation's best practitioners, available on demand for every job, regardless of the engineer's experience level.

The platform is also built with the compliance requirements of aerospace and defence customers in mind. It is ITAR-compliant, meaning it meets the International Traffic in Arms Regulations that govern the export of defence-related technology, and it can be deployed on AWS GovCloud, the secure cloud environment used by US government agencies and their contractors. These are not afterthoughts — they are prerequisites for doing business with organisations like Blue Origin, which operates under some of the strictest regulatory oversight in any industry.

From Stealth to Rocket Parts and F1 Components: Real Deployments That Validate the Vision

The most compelling evidence for Limitless Labs' approach is not the funding it has raised or the investors it has attracted — it is the customers who have chosen to run the platform in live production. Landing a paid production contract with any single aerospace or motorsport organisation would represent a meaningful milestone for a startup at this stage. Landing them with Blue Origin and Cadillac Formula One simultaneously, while also working with Sandvik and Iscar, speaks to something more substantial.

Blue Origin, Jeff Bezos's commercial spaceflight company, builds vehicles intended for repeated human spaceflight. The components that go into rockets like New Glenn cannot fail. Programming errors in CNC toolpaths do not result in scrap metal alone — they can compromise structural integrity in ways that have catastrophic downstream consequences. The tolerance requirements are extreme, the material choices are often exotic, and the certification processes are exhaustive. That Blue Origin has moved Limitless Labs into production use means the company passed an evaluation process that is not designed to be easy to pass.

Cadillac's Formula One programme presents a different but equally demanding set of challenges. F1 cars are rebuilt between races. Components are manufactured under extreme time pressure, often with lead times of days rather than weeks, in materials chosen for the best possible ratio of stiffness to weight. Programming those parts efficiently without sacrificing accuracy is exactly the kind of problem where a 50% reduction in CNC programming time translates directly into competitive advantage. In motorsport, where performance margins are measured in hundredths of a second, that kind of efficiency gain is not a marginal improvement — it is strategically significant.

For Sandvik and Iscar, both of which are deep in the business of cutting tools and manufacturing process knowledge, the collaboration has a different dimension. These are companies with decades of machining data and deep expertise in how tools perform across different materials and cutting conditions. Working with Limitless Labs suggests they see value in a platform that can systematically encode and apply that knowledge at scale, rather than relying on individual engineers to carry it in their heads.

A Competitive Landscape Where Focused Depth Beats Broad Ambition

Physical AI and industrial automation have attracted significant capital in recent years, and Limitless Labs operates in a market with well-funded competitors. Understanding where the company sits relative to those competitors helps clarify what makes its position defensible. Bright Machines, which raised a $126 million Series C in mid-2024 with backing from BlackRock, Nvidia, and Microsoft, and has accumulated over $400 million in total funding, is one of the most prominent names in industrial AI. However, Bright Machines focuses on software-defined assembly automation for electronics manufacturing — a very different workflow from CNC machining of metal parts for aerospace, defence, and motorsport. The two companies compete for investment attention and general "industrial AI" narrative space, but they are not fighting for the same production floor deployments.

Instrumental, which raised $55 million for AI-powered visual inspection technology, addresses a genuinely important manufacturing problem in defect detection and quality control, but again operates in a fundamentally different part of the manufacturing process. PhysicsX, which raised $135 million with Nvidia as a reported participant, develops physics-simulation AI for aerospace and automotive design — adjacent to manufacturing but focused on the earlier stages of the engineering workflow. None of these companies are competing directly for the CNC programming workflows that Limitless Labs has specifically targeted and built its platform around.

This specificity is both a constraint and an advantage. The addressable market for CNC programming automation is enormous — CNC machining is used in virtually every sector of precision manufacturing, from medical devices and semiconductor equipment to defence systems and consumer products. But it is a market that requires genuine domain depth to serve well. General-purpose AI tools have repeatedly struggled to translate their capabilities into reliable production performance in regulated manufacturing environments. Limitless Labs' decision to build a foundation model trained specifically on the physics of machining, rather than adapting a general model to this domain, is a bet that technical depth and domain focus will prove more durable than breadth.

Why Investors Are Looking Beyond the Screen and Into the Factory Floor

The investors who co-led this round bring strategic relevance that goes well beyond their capital. Dell Technologies Capital, the corporate venture arm of Dell Technologies, has a clear interest in how AI is deployed across enterprise infrastructure. Manufacturing is one of the most infrastructure-intensive sectors in any economy, and the intersection of AI and physical production systems represents exactly the kind of enterprise AI story that resonates with Dell's broader commercial interests. Square Peg, the Australian-born venture fund with a growing presence in Israel and Southeast Asia, has developed a reputation for backing technical founders early, often in markets that are less crowded with venture capital than Silicon Valley.

Grove Ventures, also based in Israel, is a deep-tech specialist whose portfolio has included companies that build technology at the intersection of physics and computation. General partner Lior Handelsman, who co-founded SolarEdge before becoming a venture investor, noted that Limitless Labs' achievements over the eighteen months since Grove first backed the company had exceeded expectations, specifically calling out the team's ability to combine deep technical innovation with practical software in a way that could reshape how the world's most critical parts are made. Yair Snir, managing director at Dell Technologies Capital, described the company as representing the next wave of enterprise AI — one that moves beyond digital workflows and into the physical world of precision manufacturing.

This framing matters because it reflects a broader shift in how the most sophisticated technology investors are thinking about AI's trajectory. The first era of applied AI was largely about knowledge work and software-native processes. The second era, which is now beginning, is about the physical world: factories, supply chains, industrial equipment, and the millions of workers whose expertise is embedded in physical processes rather than digital documents. The manufacturing sector alone represents trillions of dollars of global economic output, and it has barely begun to absorb the capabilities that AI now makes possible.

The new funding will be directed toward expanding US commercial operations, growing the deep-tech research laboratory in Tel Aviv, advancing the Physical AI Foundation Model toward what the company describes as closed-loop CNC automation — a future state in which the system not only programs parts but continuously learns from production outcomes to improve its recommendations over time — and roughly doubling the company's headcount over the next twelve months. With $27.3 million in total funding, a production-stage customer list that most AI startups would struggle to build in five years, and a technical foundation built specifically for the realities of physical manufacturing, Limitless Labs is making a compelling case that the most important AI story of the next decade will not unfold on a screen — it will unfold on the factory floor, one precision-machined part at a time.