Eight new and existing investors took part in the investment, which is the company’s Series D funding round, the Tokyo- and Chicago-based company said.

Among them are Moore Strategic Ventures, Coreline Ventures, Toyota’s growth-stage fund Woven Capital, and HR Tech Fund, the corporate venture arm of Japan’s Recruit Holdings. One new investor was not identified. Existing backers Atomico, Globis Capital Partners, and the JPS Growth funds, managed by a Japan Post Bank subsidiary, also took part in the funding.

The valuation is more than double the $470 million CADDi reported in March 2025. The new round brings CADDi’s total funding to $234 million, the company said.

Founded in 2017, CADDi’s initial product, called CADDi Drawer, was designed to address a common problem in manufacturing firms: they buy too many similar parts from different suppliers. The AI-powered product ingested technical drawings and then searched a customer’s own databases for similar or identical parts the customer had previously purchased or already has in inventory. The software also provided information on the defect rate of those parts, allowing the customer to decide if they wished to use existing stock, repurchase the item from an existing supplier, or try a new supplier.

In the past two years, the company has broadened its product suite, creating what it calls an “AI data platform for manufacturing.” The platform can integrate different data types from across multiple systems that customers use—from CAD files to enterprise resource planning software to HR systems—and structure it for use by both people and AI agents. CADDi Drawer has been renamed CADDi Explorer and is now joined by CADDi Agent, an AI agent designed to help manufacturing companies make decisions about standardizing parts and perform quality impact assessments, which analyze how a given design change will impact performance and safety.

CADDi has also launched six “workflow” products aimed at specific tasks, designed in part to capture the tacit knowledge of experienced engineers and workers. For instance, CADDi Design Review flags potential errors in new drawings and CAD models based on past problems with similar parts.

Yushiro Kato, CADDi’s cofounder and CEO, tells Fortune that CADDi uses its own proprietary AI model to analyze product data like drawings and CAD files, and general-purpose large language models for documents and spreadsheets. “I’ve never seen anybody who uses LLMs to do design reviews because it doesn’t understand drawings or CAD,” he said.

More than 80% of the knowledge about manufacturing work processes, and often why a company chose a particular supplier or designed a part in a particular way, is never recorded anywhere, Kato said. Instead, it exists in the heads of experienced employees. CADDi’s AI platform is designed to capture and codify that knowledge.

Kato declined to disclose revenue or customer numbers, but said sales are more than doubling year over year and that the company now has customers in 22 countries, although the U.S. is a core focus. In Japan, he said, more than half of the country’s 100 largest manufacturers use CADDi. Meanwhile, CADDi’s headcount has grown to about 900 staffers, up from 600 in early 2025. 

He said the money from CADDi’s latest fundraise will go toward expanding its product lineup, building AI models that understand manufacturing-specific data such as 3D CAD files and 2D drawings, global expansion centered on North America, and hiring.

CADDi tends to market its products based on measurable returns to its customers, such as lower direct material costs or shorter engineering lead times—critical, he said, for automakers and other manufacturing firms competing with Chinese rivals.

The biggest obstacle to adoption, Kato said, is change management. Getting workers to alter how they have traditionally done things takes hands-on help, which is why CADDi employs more than 100 customer success staff, outnumbering its salespeople. Like many AI companies, it’s started hiring “forward deployed engineers” to help customers use AI effectively. “The goal is to change the organization and create a business impact,” Kato said.

Kato frames CADDi’s ambitions around what he calls “the physical bottleneck.” AI capabilities are compounding, he said, yet little in the physical world has changed since ChatGPT debuted. Today, AI can build a e-commerce marketplace website in hours. Developing a new car, by contrast, still takes about four years from planning to delivery. “Even if AI makes thinking ten thousand times faster and produces ten thousand times the theory, the upside from AI gets diluted in the physical world if it still takes four years to mass-produce cars,” Kato wrote in a recent essay on CADDi’s website.

CADDi’s stated goal is to accelerate physical innovation tenfold by 2035—which, for a car, would mean four to five months from planning to delivery. Automakers typically go through about 20 design review cycles for a single product, Kato said, often because problems surface only at the prototype stage. CADDi wants to run more of those steps in parallel and catch problems earlier by pooling the know-how of veteran engineers into what Kato described as a kind of “superhuman” veteran.