CADDi Co. Ltd., an artificial intelligence startup that makes manufacturing equipment, said on September 16, 2026, that it raised $114 million when it comes to new funding at an estimated value of $1.2 billion, part of which will go toward expanding in North America.
CADDi develops software that enables manufacturers to unleash the power of their standard engineering drawings. CADDi Drawer, which was released in 2022, can read blueprints, and engineers can search for previously stored part designs by shape type or specifications. Manufacturing teams employ the same search function to find duplicate components and see what suppliers are charging. Today over half of the 100 largest manufacturers in Japan are customers.
The company launched a new generation of its Manufacturing AI Data Platform in Japan in August 2026 and is now starting operations in the U.S., which CADDi defines as the world’s No. 2 manufacturing market. The bulk of the data it deals with is spread throughout the business, with purchase orders, design files, and inspection findings usually staying with the team that created them.
The software interprets what the records in question mean and how they connect to each other, CADDi said. The staff and the AI agents share a single pool of resources, and there is no need to pass files around by hand.
CADDi Explorer, which is a manufacturing discovery engine, replaces Drawer, as the company describes it. Also new is CADDi Agent, which analyses the data and operates on it within the operational setting of each customer. Six workflow products encompass the manufacturing value chain.
The platform’s case was made by Yushiro Kato, who is the co-founder and chief executive, in a recent essay on the website of the company. He wrote, “Over three decades, we made the world of bits more than a hundred times faster,” while the world of atoms moved approximately at the same tempo in most industries. Kato wants to see manufacturing processes that once ran sequentially run in parallel. He says that general-purpose artificial intelligence is unable to deal with the data without domain-specific modelling.
With Fortune he was more blunt. “I’ve never seen anybody who uses LLMs to do design reviews because it doesn’t understand drawings or CAD,” Kato said this to the publication.
The lead investor in CADDi’s Series D round was not disclosed. New backers included Coreline Ventures, Moore Strategic Ventures, and Salesforce Ventures, as well as Toyota Motor Corp.’s Woven Capital, along with the HR Tech Fund, which is a corporate venture vehicle that is owned by Recruit Holdings Co. Ltd., Atomico, and Globis Capital Partners, as well as JPS Growth Investment Limited Partnership, returned from previous rounds.
The latest being a part of North American expansion was a $38M Series C extension led by Atomico in March 2025, at which point CADDi reported a $470M valuation. Since then, CADDi has increased its workforce by about half to close to 900 people, and its sales have more than doubled each year.
Says one of the partners at Coreline Ventures, Ken Hara, “AI only delivers where a data layer and a semantic layer already exist. In manufacturing that layer is buried in information fragmented across departments and in records that were never digitized at all, which is why horizontal tooling never reaches it.”
Luca Eisenstecken, who is a partner at Atomico, said CADDi has moved beyond taking care of individual workflows and is now building the intelligence layer for the overall manufacturing lifecycle.
The company also plans to invest in its own artificial intelligence technology and expand its platform development and hiring, in addition to the North American expansion.
It is worth noting that Kato had worked at McKinsey & Co. and started CADDi in 2017 with Aki Kobashi, a former employee of Apple Inc. The company, which is now headquartered in both Tokyo and Chicago, has since raised $234 million in total, including its Series D.