AI is embedded everywhere at Walmart

At Walmart, Artificial Intelligence (AI) and machine learning are everywhere.

You won’t see it when you walk down the aisle of a Walmart store. When you choose a Walmart package from your stop you will not feel it. And when you search on Walmart’s website for everything from paper towels to toys, you won’t find it.

But today, AI and ML are embedded throughout the Walmart organization – from supply chain management and shopping to discovery.

As the world’s largest retailer, it is not surprising that Bentonville, Arkansas-based retail leader has been investing in sophisticated AI for years: for example, in 2017, VentureBeat highlighted Walmart’s massive increase in inventory due to AI, while we covered Walmart. AI has been working on everything from express delivery to grocery delivery robots in the last half decade.

And over the past six years, Walmart has grown from a handful of in-house data scientists to hundreds, according to Mr. Venkatesan EVP of US Omni Tech at Walmart. These data scientists serve teams related to supply chain forecasting, optimization and labor / demand planning; Discovery and personalization; As well as emerging technologies. “We’re really spending a lot on internal development because we think this is our competitive secret sauce,” he told VentureBeat.

Venkatesan, who runs all the technology teams that enable Walmart’s global marketplace, omni supply chain and stores, said that Walmart “is evolving from being a retailer’s automator to enabling retail – that’s where AI and ML are so relevant to us.” . “What is meant by” activation “versus” automating “is that the company has taken a step back from using technology to make Walmart’s tools and processes more efficient,” he said. We want an overall end-to-end picture to enable improvement, “he said.

Which, of course, leads directly to Walmart’s biggest goal: to find and deliver what the customer wants. “Walmart always stays about what the customer wants,” he said. “Customer is number one.”

AI overwhelms Walmart’s entire supply chain

To give consumers what they want in an age of global supply chain problems, Walmart has focused on AI on the supply chain front: Last week, the company announced it would open four next-generation fulfillment centers (FCs) over the next three years. Debuted first this summer in Joliet, Illinois.

These FCs will be the first of their kind for Walmart, which will use robotics and machine learning to accelerate fulfillment. The company claims that together with its traditional fulfillment centers, Walmart will now be able to reach 95% of the US population with next- or two-day shipping.

In addition, Walmart announced last Monday that it would roll out Symbotic’s next-generation robotics and AI technology to all 42 of its regional distribution centers over the next eight years – already in its 20s – as the retailer works to modernize its supply chain network. The company said in a statement that the technology should help Walmart increase the accuracy of its inventory and help its warehouses increase their ability to receive and ship products to stores.

Symbotic, which went public this week and enjoys significant investment from Walmart, said its AI-powered software and robotics system – including its symbols – are fully autonomous vehicles that take advantage of machine learning, vision and algorithms. Some of the biggest challenges are addressed. Walmart’s complex supply chain.

“When you look at things like accuracy and reduced errors and reduced scrap, there are only incredible savings from working capital perspectives, inventory management perspectives and overall labor pieces,” said Michael Loparko, CEO of Symbotic. “So I think there are powerful cost drivers – but I think the biggest catalyst for Walmart is changing customer demand and the need to pull the market.”

Walmart’s developed supply chain

Walmart’s supply chain efforts using AI have evolved over the past few years, Venkatesan said, to predict customer demand in terms of what the customer really wants to buy – to predict sales demand – how much will sell already in stores – from Google search. By analyzing data on channels ranging from tick tock to social feeds.

During the epidemic, however, the difficult demand problem to solve also became a supply thorn problem.

“We learned that we need to understand what is not in stock and what we should change,” he said. “So we put a lot of effort into AI and ML for substitution logic.” Deep Learning AI takes into account hundreds of variables – size, type, brand, price, overall shopper data, personal preference and current inventory, among others – in real time to determine the best next-available item.

AI powers Walmart’s discovery and personalization

Historically, much of Walmart’s activity around discovery and personalization has revolved around automated decision-making, said Jan Pedersen, VP of Search and Personalization at US Omni Tech at Walmart. But more recently, the performance of the Computer Vision AI model has become much better than before because of deep learning, he explained. “You can use these items in production and get results,” he said.

As a result, there are many areas where Walmart uses AI techniques and natural language processing in search and personalization, he explained. There are English language queries – to understand what people mean when they request a product type, to understand which parts of the query are important.

Understanding the quality of the image is also key, he added. “Maybe you’re also doing attribute extraction, so know it’s a red shirt because it’s red in the picture.” Finally, there is machine translation. “We don’t need to translate anything manually, so that’s a big incentive,” he said.

Search is an expanding boundary

Some questions, however, are easier than others, he pointed out. “You may have a query that is repeated many times and people give you a very strong indication of what it means, or you may have a query where if you see it the purpose is clear to the user. What’s the point, but if you attack it with a standard approach, you won’t get really good results. ”

A recent example of this, he explained, is ‘avocados from Mexico.’ “The reason it’s interesting is that most avocados don’t tell you they’re from Mexico.” On the other hand, he explained, the query itself is very clear – it is clear what the user wants. “So we put it in a bucket of semantic queries where you really should be on top of it, realizing that the avocado portion is important or generally guess from other things that you know about the item that might be important.”

Finally, Pedersen discussed Walmart’s efforts regarding multilingual questions, enabling Spanish-speaking customers to find specific items on the site and in the app.

“One of the interesting things about search experiences in general is that people can write whatever they like, because it’s an empty box,” he said. To serve Spanish-searching customers, Walmart uses language search using AI. “You find that the query is likely to be in Spanish and then you machine translate the query into English,” he said. “Then, when we get the result, we return the result in English. The next step is to machine translate the product description material so that we can translate the titles. ”

AI-powered fitting room tech

Computer Vision also powers one of Walmart’s most recent AI-powered offerings: Dynamic Virtual Fitting Room technology from Zeekit, which Walmart acquired last year. It allows customers to buy clothes online and see what an item really looks like on them.

Walmart’s “Choose My Model” experience, which launched in March, offers customers a choice of 50 models between 5’2 “- 6’0” height and XS – XXXL size. Customers can choose the model that best represents their height, body shape and skin tone.

“Based on the millions of images we have from the catalog, we analyze all the different issues on the models and use them to create dress simulations,” said Desiree Gosby, VP of Walmart’s New Business and Emerging Tech. “Whether it’s going to fit loose, where the waist is going to fit, it’s about breaking down everything about how the length should be adjusted depending on your height.”

Currently, Gosby’s team is working on an experience using Zeekit’s technology where customers can actually upload their own photos. “It’s actually a difficult issue for AI and Computer Vision,” she said. “And customers have to make sure they’re taking a good picture that they feel good about.”

Conversational AI in the mix

Walmart also recently launched, after several months of testing, its conversation known as AI technology text to shop. Customers can text or say whatever they want and Text to Shop will add it to their cart. If they need an item they have never purchased before, Walmart will offer product recommendations

“It’s really about how we make it easy for the customer to express what they want or what they want from us,” Gosby said. “It’s basically a digital support platform that takes advantage of voice and text chat – we work with the entire company, including customer care, and we power Walmart shopping assistants in Google and Siri.”

Text to shop is the result of a lot of investment in understanding natural language, she added. “We’re taking advantage of GPT-3 under the hood and then really taking advantage of our data to create a natural language understanding that’s natural.”

But, she admits, “it’s really hard to make it easy – if you say things like adding chocolate milk and pizza to my cart, be able to understand that you mean chocolate versus milk versus chocolate milk.”

Overall, these techniques are about giving consumers the confidence to make a purchase, Gosby said. “Are we really saving their time? Are we reducing return rates for wear? “Everything Walmart does should be about somehow eliminating friction for the customer,” she said: “We don’t do technology for the sake of technology.”

Walmart’s AI is strictly customer focused

When asked about the future of AI at Walmart, Venkatesan returned to focus on the customer. “Our forecast for the future has always been what the customer wants – we monitor the customer very carefully,” he said. “We can understand how the customer’s trends are going and then we will adapt to it, because it is very difficult to predict exactly where it will go.”

Walmart will continue refining, he added. “I think there are still a lot of improvements to be made,” he said. “What we are constantly doing will be constantly evolving or upgrading, as it will become more complex according to customer demand.”

Venturebeat’s mission Digital Town Square is set to become a place for technical decision makers to gain knowledge about the changing enterprise technology and practices. Learn more about membership.

Similar Posts

Leave a Reply

Your email address will not be published.