As factories return home and reshoring gathers pace, a hard truth is resurfacing alongside production lines: machines can be shipped, but know-how cannot. For decades, American manufacturing quietly outsourced not just labor, but institutional memory. Now, with geopolitical pressure mounting and experienced workers retiring, that loss is becoming painfully visible on shop floors.

Dirac, a manufacturing software company, believes it has found a way to rebuild that missing layer. Earlier this year, the company announced the general availability of BuildOS, which it calls the first automated platform for creating and managing model-based work instructions. The launch coincides with a surge of momentum for Dirac, including $10.7 million in funding and a strategic partnership with Siemens aimed at accelerating digital transformation across global manufacturing.

“Dirac has built the first and only automated work instruction platform,” said Tomás Klausing, Director for Technology Partnerships at Siemens, underscoring how central the technology could become to modern production environments.

Manufacturing’s Quiet Crisis: Lost Context

Across aerospace, defense, shipbuilding, and heavy industry, frontline operations still depend heavily on static PDFs, outdated SOPs, and informal shadowing to train workers and execute tasks. The most critical details—how to torque a joint, how to adjust a setup mid-process, what sequence to follow when something goes wrong—often live only in the minds of senior technicians.

As those workers retire and production demands increase, manufacturers face a bottleneck that can’t be solved by adding capacity alone. Engineering files may exist, but without context, they are unusable on the floor without weeks or months of onboarding.

Dirac CEO and founder Filip Aronshtein frames this as a structural problem, not a staffing one. “We don’t have a labor crisis; we have a context crisis,” he said. “You can’t reshore production if nobody remembers how to build.”

The stakes extend well beyond efficiency. While the U.S. produces roughly 1.5 submarines per year, China produces more than 10 per month. In that context, modernizing manufacturing processes is not just an economic concern—it’s a national security issue.

From Engineering Intent to Repeatable Execution

BuildOS is designed to replace static documentation with automated, animated, physics-aware, and interactive work instructions. Instead of manually creating step-by-step guides, the platform automatically translates CAD files and bills of materials into visual, shareable instructions complete with 3D models, tools, and embedded tribal knowledge.

In effect, Dirac is trying to do for manufacturing engineers what CAD once did for mechanical engineers: give them a way to encode processes once and scale them reliably across teams, facilities, and generations of workers.

“BuildOS turns engineering intent into repeatable execution and tribal knowledge into institutional knowledge—automatically,” Aronshtein said.

The platform’s impact is already visible among early adopters. Ancra Aircraft, which supplies cargo loading systems to more than 60% of the global freighter fleet, reported that work instruction creation time dropped from three days to two hours, a 95% reduction. The company also saw fewer quality errors thanks to tighter process standardization, while capturing tacit knowledge from senior technicians in a single session.

Why AI Alone Isn’t Enough

Dirac is careful not to position BuildOS as just another AI tool layered onto existing workflows. Instead, the company argues that intelligence must be embedded directly into how work is defined and executed.

Artificial intelligence helps automate translation from design to instruction, but the real value lies in context preservation—making sure every worker understands not just what to do, but why and how it fits into the broader process. That context becomes increasingly critical as factories move toward low-volume, high-mix production with rising compliance requirements.

“America is investing billions in industrial capacity,” said Trae Stephens, Partner at Founders Fund, “but without platforms like Dirac’s BuildOS, we’re just recreating the same bottlenecks that made offshoring attractive in the first place.”

Stephens added that Dirac’s advantage lies in helping manufacturers compete on intelligence, not incentives. “They’re re-architecting American manufacturing from the ground up.”

Built for the Shop Floor, Not the Slide Deck

Part of Dirac’s credibility comes from its team. The company’s founders and engineers come from hardware-heavy industries—planes, cars, submarines—where the cost of miscommunication is measured in delays, defects, and safety risks. BuildOS wasn’t designed for a theoretical factory of the future, but for the realities of today’s shop floors.

That practical focus has helped drive adoption across aerospace, defense, automotive, agriculture, and heavy equipment manufacturing, particularly among companies managing complex products with small batch sizes.

Dirac and Siemens: Closing the Digital Loop

The partnership with Siemens adds another layer to Dirac’s ambitions. Together, the companies aim to close the gap between product design and physical production, creating a tighter feedback loop between PLM and CAD systems on one side and ERP and MES platforms on the other.

“The partnership with Dirac enables us to provide additional efficiency to our customers,” Klausing said. Early integrations have already improved work instruction generation, manufacturability, and production planning. Siemens plans to deepen the integration as part of its broader digital manufacturing portfolio.

For Siemens, the collaboration supports a strategic goal: bringing emerging technologies to market faster through flexible partnerships rather than monolithic platforms.

Toward Smart Labor, Not Cheap Labor

As manufacturing shifts from mass labor to smart labor, BuildOS positions itself as a bridge between generations of workers. The platform allows companies to capture the expertise of experienced operators and make it immediately available to new hires—reducing ramp-up time while maintaining consistency.

“Our edge isn’t just speed,” Aronshtein said. “It’s repeatability. BuildOS gives every engineer, technician, and team the ability to build something once, and then build it right every time after that.”

In an era defined by reshoring, workforce transition, and geopolitical pressure, Dirac is betting that the most valuable asset a factory can protect isn’t machinery—it’s memory. And as American manufacturing attempts to relight an industrial engine long deprived of its wiring diagram, platforms like BuildOS may determine whether that engine runs smoothly or stalls again.

Get in touch with our MarTech Experts.