Ex-OpenAI Exec's Arda Hits $1.5B on Factory AI

Bob McGrew, OpenAI’s former research chief, has raised $230 million for Arda, an AI manufacturing software startup valued at $1.5 billion. The company told CNBC in an exclusive interview Tuesday. Altimeter Capital, Accel and Khosla Ventures led the round. It’s a striking vote of confidence in a company aiming to automate the factory floor.

Arda builds AI software to help automate manufacturing. McGrew said its models are already deployed with military technology company Anduril. That’s notable for a startup still early in its commercial life, and it points to demand from the defense sector, where production timelines matter most.

The Product

McGrew spent more than a decade as Palantir’s product and engineering lead. He described Arda’s offering in plain terms: “a user interface and an agent that does for manufacturing engineers what Claude or Cursor does for programmers.” The comparison is deliberate. Software engineers have seen AI tools compress their workflows, and McGrew is betting that the same logic applies to factory design.

The bottleneck, he argued, is human time. Manufacturing is “human-intensive,” and adapting production to a new weapon or drone design can take considerable effort. “When a design changes, you can retrain a human in a day — but it takes three to six months in order to change the automation to reflect the new design,” McGrew said. “We’ve built a software and AI product that basically allows manufacturing engineers to build assembly lines much faster.”

That gap is Arda’s opportunity. Retraining a worker is faster than reprogramming a line. It’s a practical problem, not a theoretical one, and the kind of friction private firms tend to solve faster than large bureaucracies.

A Founder’s Path

McGrew joined OpenAI in 2017 and worked on the reasoning capabilities of early models before ChatGPT launched. He left during a 2024 wave of high-profile departures that included chief technology officer Mira Murati and co-founder and chief scientist Ilya Sutskever. “OpenAI is an amazing place,” he said.

He recalled sitting in rooms with CEO Sam Altman, Dario Amodei, who left to start Anthropic, and Sutskever, “talking about what it would look like when a model could write your email for you.” The anecdote is small, but it shows how quickly the field’s ambitions have grown from consumer conveniences to industrial infrastructure.

McGrew also stayed through the weekend in November 2023 when Altman was abruptly fired and rehired. He said he wanted to stay to ship OpenAI’s reasoning product, which he called the “last key breakthrough left before AGI,” or artificial general intelligence. “I said, if I could just ship that, I’d be happy with my career at OpenAI, and I’d be ready to move on to something else,” he said. He did exactly that.

Market Impact

Arda runs its software on a mix of proprietary models, open-source models and some frontier models. The company’s other co-founders are Augustus Odena, who helped start Adept and previously worked at Google and Meta, and former Palantir engineers Alex Mark and Jakob Frick. The team brings extensive experience building and scaling software companies, not just research labs.

Altimeter’s Apoorv Agrawal led the funding round after meeting McGrew when both worked at Palantir. Agrawal framed the opportunity in stark numbers. “Manufacturing globally is a $16 trillion industry. In the U.S. alone, it’s $3 trillion,” he said. “We think that Arda, and the vision that they’ve outlined, will continue to capture a significant portion of that. Yes, the price is steep today, but the long-term opportunity is very high.”

Agrawal acknowledged that manufacturing largely happens overseas, but said there’s a market for bringing it back. Investors are wagering that private capital and better software can make domestic production competitive without subsidies doing the heavy lifting. That’s a market bet, and results will decide it.

Reshoring and the Roadmap

McGrew said robotics is next on Arda’s product roadmap. He argued that factory automation could help bring manufacturing and jobs back to the United States. “If your design changes very frequently, the only place you can really build your product is in East Asia, because you need an army of human workers who can be retrained quickly, and you can’t wait for automation to catch up,” he said.

He also cited the Trump administration’s effort to onshore manufacturing. “No one has ever successfully built a smartphone in the U.S.,” McGrew said. “Never successfully assembled one, and so if you can build a smartphone in the U.S. that is not taking away American jobs. That is bringing American jobs.” It’s a bold claim, and it will test whether software can close a cost and speed gap built up over decades.

Why This Matters:

Arda’s $230 million raise at a $1.5 billion valuation shows private investors still see real money in industrial software, even as the price tag runs high. Agrawal conceded as much, saying the price is steep today. For taxpayers and policymakers focused on reshoring, the question is whether a private company can deliver faster assembly lines than the government’s incentives have so far produced. Arda’s early deployment with Anduril suggests defense buyers are already willing to pay for speed. If the company delivers, gains will come through private markets, competitive pressure and worker retraining rather than new federal programs. If it doesn’t, investors will absorb the losses, not the public. Either way, the outcome will say something about how much faster a market-driven approach can move than government-led industrial policy.