Manufacturing Voices: Yushiro Kato, Co-Founder and CEO, CADDi

Read the article in Design World

By Sana Kazilbash | April 2, 2026

This article is sponsored by CADDi. In this Voices interview, Design World spoke with Yushiro Kato, co-founder and CEO at CADDi, to discuss his founder journey and why unlocking historical engineering knowledge is the key to driving more effective Value Analysis and Value Engineering (VA/VE) initiatives in American manufacturing.

Design World: Tell us about yourself and your role as co-founder and CEO of CADDi.

Yushiro Kato: I founded CADDi at the end of 2017 with my co-founder, Aki Kobashi, who’s the CTO at CADDi. I used to work for McKinsey, a management consulting firm, for three and a half years as an engagement manager, co-leading procurement for the manufacturing sector and spending plenty of time on the shop floor helping expand the firm’s IoT and AI practice. I moved to Chicago almost three years ago to lead the U.S. market expansion of CADDi, and I’m currently the global CEO as well. I call it “Chief Everything Officer”.

You worked with American manufacturers at McKinsey and launched CADDi first in Japan. How have those experiences shaped CADDi’s platform and advisory approach for the U.S. market today?

I had my own business when I was at university before McKinsey. Then I joined McKinsey to learn more about the issues in traditional industries like manufacturing. I didn’t specifically target manufacturing at the time, but I wanted to do something bigger.

I did a lot of projects at McKinsey, including manufacturing, construction, and investing firms, among others. Manufacturing was the most fascinating industry for me because I had three criteria at the time for the industries that I wanted to target. One was the size of the industry, as large as possible, and second the depth of pain points. I didn’t want to do something that I could solve within two or three years with simple technology and action. And then three, how common the industry was globally. Those are the three criteria I had, and manufacturing was the best match.

I spent about two years at McKinsey focusing on manufacturing, especially supply chain transformation. This included procurement cost reduction, new supplier development and design improvement for manufacturability, and cost reduction — which is called value analysis and value engineering (VA/VE). I worked with some U.S. manufacturers, large enterprise companies in Milwaukee, Nebraska, Kansas City, and New Jersey. Honestly, I saw a lot of commonalities across Japan and the U.S. in terms of the pain points they have and the solutions that can impact those pain points.

We actually started with a business called CADDi Manufacturing, which is a procurement platform for custom-made parts. We were not selling software. We were selling physical parts like Amazon. Behind it, we had implemented a lot of software to streamline our own operations because we had 5,000 customers on the manufacturing platform and 600 suppliers. Every company has its own format and its own strengths and weaknesses. We couldn’t handle it manually. Naturally, we implemented software, and now we are selling the software we used internally.

I learned about a lot of pain points, especially in the supply chain area at McKinsey, and I leveraged that foundation to establish CADDi.

Manufacturers generate decades of engineering knowledge through drawings, supplier decisions and product revisions. Why do you think so much of that knowledge remains underutilized inside organizations?

We did CADDi Manufacturing for seven years from the foundation and became one of the largest on-demand manufacturing platforms in the world. What we learned at the time was that we owned more than ten physical inspection centers to assure the quality of the parts, and we delivered the parts to customers. Like Amazon, we had a lot of warehouses and inspection centers.

One purpose was to track quality issues, but the second purpose was to gain data to identify what causes quality defects. We identified that more than 50% of quality defects are caused by the lack of quality in drawings and designs. That’s why we developed CADDi Drawer and launched it four years ago to change that.

Engineers have their own systems, like product lifecycle management (PLM) systems and drawing management systems. They design based on past designs and their tribal knowledge. But to prevent defects, you need to understand what kind of defects can happen if you change this design to another design, or if you tighten a tolerance. The systems are siloed. Procurement people use their own supply chain management systems. Quality people use their own quality management systems, QMS or Excel files, and track past quality reports in PDFs. It’s all separated and disconnected.

To change quality defects, you need to change designs, but you don’t have that data or visibility. That’s one of the biggest reasons why transformation is hard to achieve.

What we did was connect all the data from siloed systems. We also developed our own proprietary shape analysis AI model to understand the similarity across parts, even from 2D drawings. A human brain can understand that these parts are similar to another part, so we can leverage the learnings from that similar part in the past for the part I’m working on. But AI or software couldn’t do that. So we developed it to connect different systems through part number or part shape.

That’s our innovation. That’s the storyline of why it was not easy to solve and now why we can solve it.

At what point did you realize that unlocking historical engineering knowledge could influence not only engineering productivity but also sourcing, supply chain decisions and overall competitiveness?

I realized the power of engineering because we were a supplier, a platform for engineers and procurement buyers. More than 50% of the quality defects happened because of the lack of quality of engineering and drawings.

Customers always asked us to reduce costs when we were doing CADDi Manufacturing. Oftentimes, if you want to reduce costs, there are some commercial levers to negotiate with suppliers, but you can realize more impact through changing designs. If you change materials, it’s the easiest way to reduce costs, for example. Or if you change the tolerance, if you can loosen one tolerance, you can gain cost reduction impact.

We learned that through millions of parts transactions at CADDi Manufacturing. After we launched CADDi Drawer, all the data was connected and a lot of engineers used it to learn past quality defects and cost reduction histories of similar parts using the similarity search that we have. A lot of customers, like Subaru, an automotive company, realized the cost reduction impact very quickly.

The CADDi Manufacturing experience helped us realize that engineering is super-important and it has a lot of impact on the downstream funnel in terms of cost reduction, lead time and more, through inspections. Then through the experience of CADDi Drawer and software implementation, we learned that this impact can be realized in real life with specific customers.

Many engineering teams struggle to execute cost-down initiatives like Value Analysis and Value Engineering (VA/VE) because they lack access to historical design data. How do you expect manufacturers that successfully unlock this knowledge to operate differently when it comes to VA/VE and continuous innovation?

If you are a senior engineer, you have a lot of experience. If you change this material from A material to B material, you will realize this cost reduction. Or if you change this feature to that feature, you can realize this. This is what VA/VE is, changing designs to reduce cost and reduce quality defects.

But for most younger engineers, they don’t have that tribal knowledge. They don’t have that much experience. So it’s hard to come up with VA/VE ideas.

What we are doing is creating visibility across similar parts from the past, regardless of the designers. Even if you haven’t designed similar parts before, we can identify similar previous parts and what kind of quality defects happened, or what kind of design reviews and design changes happened to those drawings. Then they can learn from past experiences rather than recreating similar parts without understanding the historical context.

That’s what we are doing. Basically, AI identifies similar parts and similar VA/VE cost reduction history and quality defects to give insights to younger engineers for VA/VE initiatives.

Looking ahead, how do you see AI changing the way manufacturers use historical engineering knowledge to make faster and more informed decisions?

The concept of our product is to prevent reinvention of the wheel, which is happening everywhere in manufacturing. It is said 99% of the activities manufacturing workers are doing are just reinventing the wheel to some extent. Someone in your company has likely done similar things in the past. Oftentimes a lot of knowledge is only in human brains. It’s not tacit knowledge, it’s just tribal knowledge. If people leave the company, that knowledge will also walk out.

What we are doing is to assetize, to capitalize on people’s knowledge, past data and experiences as a company-wide asset like a knowledge base, so that you don’t have to reinvent the wheel. Instead, you can spend your time creating new things, innovative stuff that no one else in the company has ever worked on. That’s what we want to achieve, and AI is the best tool to do that.

We do not just leverage large language models (LLMs) that are available in the market. The biggest advantage we have is the shape analysis model that we created based on the proprietary data we accumulated through CADDi Manufacturing and real manufacturing experiences. This is the biggest difference between CADDi and normal system vendors, because we were a manufacturer. I think the most important asset is the human experiences, historical data, and failures in the company—and AI can unlock the potential of all that historical data and leverage it as an asset.

CADDi is a global technology company that develops a manufacturing-exclusive data intelligence platform. Headquartered in Tokyo and Chicago, CADDi brings proven expertise in helping American Manufacturers preserve decades of engineering knowledge. Its flagship product, CADDi Drawer, uses advanced AI to structure fragmented engineering data into searchable intelligence. Recognized globally for innovation, CADDi was listed in Fast Company’s Most Innovative Companies and received the SaaS Award for Best Business Intelligence and Engineering Management Software.

Learn more in CADDi’s whitepaper, Value Analysis & Value Engineering for Manufacturing: The Challenges and Solutions of Implementation.