What Happened

CADDi, a Tokyo- and Chicago-based manufacturing AI startup, has raised $114 million in a Series D funding round that values the company at $1.2 billion, according to an exclusive Fortune report published September 15, 2026. The valuation is more than double the $470 million the company reported in March 2025.

The round included eight new and existing investors: Moore Strategic Ventures, Coreline Ventures, Toyota's growth-stage fund Woven Capital, Salesforce Ventures, and HR Tech Fund (the corporate venture arm of Japan's Recruit Holdings). Existing backers Atomico, Globis Capital Partners, and JPS Growth funds also participated. Total funding to date is $234 million.

Founded in 2017, CADDi started with a product called CADDi Drawer — now renamed CADDi Explorer — that ingested technical drawings and searched a manufacturer's databases for similar or identical parts to reduce redundant purchasing. In the past two years, the company has expanded into a broader "AI data platform for manufacturing" that integrates data from CAD files, ERP systems, and HR systems, and structures it for use by both people and AI agents.

The company now offers CADDi Agent (an AI agent for parts standardization and quality impact assessments) and six workflow products targeting specific manufacturing tasks, such as CADDi Design Review, which flags potential errors in new drawings based on past problems with similar parts.

CEO Yushiro Kato told Fortune that CADDi uses proprietary AI models to analyze drawings and CAD files, and general-purpose LLMs for documents and spreadsheets. He noted that more than 80% of manufacturing process knowledge is never recorded — it exists only in the heads of experienced employees. CADDi's 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. The company has roughly 900 employees, up from 600 in early 2025, and operates in 22 countries. In Japan, more than half of the country's 100 largest manufacturers use CADDi.

Why It Matters

CADDi is addressing a problem that most AI companies have ignored: the physical world hasn't changed much despite AI's rapid compounding capabilities. Kato frames this as "the physical bottleneck" — the gap between what AI can do in software and what actually changes on factory floors.

The technical moat is real. General-purpose LLMs cannot parse 2D engineering drawings or 3D CAD files. CADDi's proprietary models can, which means the company isn't easily displaceable by a GPT upgrade. This is a critical lesson for AI startups: vertical AI defensibility often lives in proprietary data formats that foundation models can't handle out of the box.

The investor lineup is strategically significant. Toyota's Woven Capital brings automotive manufacturing credibility and potential customer access. Salesforce Ventures signals enterprise software ecosystem alignment. The $1.2B valuation on $234M total funding — a roughly 5x valuation-to-total-funding ratio — indicates investors are pricing in strong unit economics and growth, not just hype.

Perhaps most instructive is CADDi's go-to-market model. The company employs more than 100 customer success staff, outnumbering its salespeople, and has started hiring forward-deployed engineers. This reflects a hard truth about AI adoption in manufacturing: the technology is not the bottleneck — change management is. Workers have done things the same way for decades, and getting them to adopt AI-assisted workflows requires hands-on support, not a self-serve SaaS funnel.

Who Is Affected

Manufacturing companies — especially automotive, industrial equipment, and electronics manufacturers — who are evaluating AI tools for engineering data management, parts standardization, and quality impact assessment. CADDi's expansion from parts deduplication into a full AI data platform means it's now competing with both point solutions and broader enterprise data platforms.

AI startups building vertical products for industrial use cases should study CADDi's multi-model architecture and heavy customer success model as a go-to-market blueprint. Enterprise IT buyers at large manufacturers need to assess whether CADDi's workflow products can integrate with their existing CAD, ERP, and HR systems — the company claims it can, but integration depth will vary by customer.

Strategic Implications

For AI startup founders: CADDi's $1.2B valuation on $234M total funding demonstrates that vertical AI in manufacturing can command premium valuations without consumer-scale growth. The key is solving a data format problem that general-purpose models can't. If you're building vertical AI, identify the proprietary data formats in your industry that LLMs can't parse and build models for those specifically. That's your moat.

For developers/operators building with AI APIs: CADDi's architecture — proprietary models for drawings and CAD, general LLMs for documents and spreadsheets — is a practical blueprint for multi-model systems. Don't force one model to do everything. Route by data type and use fine-tuned or proprietary models where general models fail. This is especially relevant for any AI product touching engineering, medical imaging, or other domains with specialized file formats.

For non-technical business owners evaluating AI tools: CADDi's investment of 100+ customer success staff (outnumbering sales) tells you that AI adoption in manufacturing is a change management problem, not a technology problem. When evaluating AI vendors for industrial use cases, prioritize those that offer hands-on implementation support and forward-deployed engineers over those pitching self-serve platforms. The ROI depends on whether your workers actually use the tool.

What to Watch Next

Monitor CADDi's North American expansion progress — the company's U.S. customer base and revenue mix will determine whether its Japanese manufacturing playbook translates to American automakers and industrial firms. Also watch for whether competitors in the manufacturing AI space (or horizontal players like Siemens or Dassault Systèmes) respond with similar AI agent products for engineering data.

Frequently Asked Questions

Q: What does CADDi do?

A: CADDi builds AI software for manufacturers that organizes engineering and production data — including 2D drawings, 3D CAD files, ERP data, and HR data — and uses proprietary AI models and AI agents to help companies standardize parts, perform quality impact assessments, and capture tacit knowledge from experienced workers.

Q: How much did CADDi raise and at what valuation?

A: CADDi raised $114 million in a Series D funding round at a $1.2 billion valuation, more than double its previous $470 million valuation from March 2025. Total funding to date is $234 million. Investors include Toyota's Woven Capital, Salesforce Ventures, Atomico, and Moore Strategic Ventures.

Q: Why is CADDi's approach different from general AI tools?

A: CADDi uses proprietary AI models that can parse engineering drawings and 3D CAD files — data formats that general-purpose LLMs cannot understand. It combines these proprietary models with general LLMs for text-based documents, creating a multi-model system tailored to manufacturing-specific data.