Brooks Brothers is the latest fashion brand to go big on artificial intelligence. The American tailoring company has partnered with ORS Group, a company specializing in artificial intelligence, to create a platform that will integrate AI data analytics into nearly every aspect of Brooks Brothers’ supply chain.

The new initiative, despite being vast in scope, is less flashy than some other artificial intelligence uses from the fashion world. Compared to chatbots and voice assistants, which are consumer-facing but ultimately not game-changing, the backend AI tools are less visible and exciting, but they’re providing more immediate and valuable solutions.

We spoke with Brooks Brothers’ COO Gianluca Tanzi about the brand’s big bet on artificial intelligence, the ways AI has affected the brand’s business model and his boundless confidence in AI as a tool.

Right now, is AI a risk or is it a completely safe bet?
It’s a fantastic bet right now. It’s the best thing we can do. When you make decisions, you are doing so based on opinion, and opinions are not always driven by fact. AI helps you form opinions based on fact, which lead to better decisions being made. You don’t need to guess, you don’t need to speculate, you can just look at the numbers.

It’s like a pilot flying a plane. There are so many things he needs to keep track of at any moment, but he has a co-pilot who reminds him of the four or five most important things to keep track of at any moment. It’s the same with us. We have thousands and thousands of data points we could look at, and the challenge is picking out the most relevant parts. That’s where the artificial intelligence comes in. It’s our co-pilot.

Were there any aspects of your business that you were surprised AI was able to help?
Price elasticity, for sure. Everyone thinks that if you reduce price, you increase volume, but this is not completely true. It depends on the different channels of trade. Different stores have different elasticity. It depends on the population who shops in that store. Here in New York, people may be looking for more [in-store] service, maybe they are not looking only at price. At other places people might be looking for something more specifically with price in mind. Certain categories are more elastic than others. Shirts are very elastic, but outerwear is not, for example.

AI has also helped us analyze what people are buying in certain geographical regions. So take an area around the store and see what people are buying in that zone, and compare that to what you have in that store. If you do this, you discover some things that can be really strange. For example, in one store in Florida, a lot of our customers were buying shirts and ties. Now, normally, you buy a shirt before a tie. But 75 percent of people at this store bought a tie first and then looked for a shirt to match it. It’s fascinating. Then you can take a deep dive and look at why. You speak to the people in the store, you double check the data, and then you are able to connect the dots and dig up the root of the information. You find things that you never would have thought of or discovered on your own. That’s the beauty of artificial intelligence.

When were you convinced that AI was important?
It started when I met Pierluigi [Riva, founder of ORS Group and architect of Brooks Brothers’ AI platform]. We were making fabrics at Marzotto textiles. We had this department, which was fully automatic. The material was coming in and being put in an automated machine, and all the dying was automated. Lead time was very long, four weeks. We had 30 workers there. Not efficient. Pierluigi sat down and asked how can we make this more efficient. At that time, the scheduling was done by hand, by a human. We substituted that human with AI, and it went from four weeks to two, 30 workers to 20. The payback was immediate.

It seems low-risk and high-reward.
There are so many things you can do with it — like priority. Priority meaning I only have a limited amount of stock and many different stores to send it to. So what do you do? Run a simulation with certain parameters to decide the most efficient allocation of resources and see the result in 20 minutes. If you don’t like it, change the parameters, run it again, grab a coffee and check it again. It allows you to be so efficient in a way you couldn’t do with just the human mind. If I produce something in Italy that needs to ship to Atlanta, the AI can help me set it up so it gets sent straight to the store, and it doesn’t have to go to the warehouse. That cuts off three or four days from the shipping process. The efficiency ripples out in both directions. You cannot do that without a machine.

So, what are AI’s limits?
Obviously there’s are limits. A designer can use AI to support his job, but the AI can’t say that next season the predominant color will be green. It’s better to use artificial intelligence to manage more structural data. Imagine a chef. No matter what machines or intelligences he has at his disposal to help make decisions, only the chef can really taste all the flavors and make those final decisions.