Sourcing Journal

Amazon v. Walmart: Two Industry Behemoths Battle Over AI

By Meghan Hall,

11 hours ago

https://img.particlenews.com/image.php?url=0LXZZw_0uszSiui00

Many companies in the retail, supply chain and logistics industries have considered how to integrate artificial intelligence into their processes—whether through automation , predictive analytics, machine learning or generative AI systems.

While some have mature levels of adoption in certain areas, it’s difficult to argue that many companies are as far along in their respective strategies as two industry titans: Amazon and Walmart .

Some of the ways the two behemoths try to edge each other out remain largely similar, but in some
instances, the giants take different approaches to their technology development and implementation.

Amazon: Review summarization

Amazon released further information about the ways it uses AI for customer reviews earlier this year, noting that systems confirm whether a review is potentially fake.

It also stated that it uses AI to aggregate reviews and create summaries that a spokesperson said are meant to “provide customers with common themes from dozens, hundreds or even thousands of reviews at a glance to help them quickly understand customer insights.”

With the AI features as a guiding hand, customers are expected to make quicker decisions while receiving answers to FAQs and having access to aggregate sentiments.

Amazon: AI for product discovery and search

Amazon famously announced Rufus , its generative AI shopping assistant, earlier this year. According to a spokesperson, “Rufus [is] trained on Amazon’s product catalogue, customer reviews, community Q&As and information from across the web to answer customer questions on a variety of shopping needs
and products, provide comparisons and make recommendations based on conversational context.”

The company also leverages generative AI to aid sellers in writing product descriptions that will play
well with customers. The tool works to ensure that, when a seller lists a new item, its titles and descriptions catch consumers’ eye and answer their questions. Sellers can also enrich existing listings, which makes the discovery process easier for the consumer.

Amazon: Inventory management

Amazon has a system called the Supply Chain Optimization Technology (SCOT), which a spokesperson
described as “the central nervous system of Amazon’s operations.”

The spokesperson explained that the underlying systems behind SCOT make it possible for the e-commerce behemoth to manage a supply chain that has millions of sellers in it, while also
allowing those sellers to manage their own inventories and storefronts.

“It’s actually many AI systems operating in concert, and together they make countless predictions and
decisions every day. SCOT transforms our massive data sets into predictive intelligence, telling us what inventory to buy, where to store it, how to pick it and ship it. It both forecasts demand and coordinates warehouse operations. It’s like a master conductor leading an orchestra of millions,” the spokesperson
told Sourcing Journal.

Amazon: Warehouse automation

An Amazon spokesperson said that, while many of its forward-facing AI systems have already started to prove their value and improve customer experiences, the internal AI systems—some of which have been in place for many years—are the ones driving more complete transformation.

“Amazon has one of the vastest networks of operations in the history of commerce. We pick, pack, ship, sort and deliver billions of packages across the world every year. And we are constantly striving to get faster and provide the widest selection of items to customers affordably. In today’s complex supply
chain, you just can’t do that without some of the most advanced AI systems in the world,” the spokesperson said.

The company, the spokesperson shared, uses more than 750,000 robots in tandem with its human employees to fulfill orders. But the spokesperson noted that, even as it increases its fleet of robots, the company continues to add new roles to keep up with that growth.

For instance, they said, Amazon has had to train some employees as robotics floor monitors and technicians; it has also hired reliability maintenance engineers to keep the fleets running smoothly.

“By taking on tasks that are strenuous or repetitive, our robots help make work safer for our employees,” the spokesperson said.

Amazon: Last-mile optimization

For Amazon, last-mile optimization begins before the customer journey even starts, a spokesperson said.

“Delivery optimization starts long before a customer places an order. In fact, we’re using AI to predict demand, inventory distribution [and] even the number of drivers we’ll need more than three months before a customer clicks ‘Buy Now.’ For that kind of predictive intelligence you need two things: massive amounts of highly reliable data and powerful AI models to translate that data into insight,” the spokesperson said.

The company uses large language models (LLMs), neural networks and ML to ensure drivers are able to take the safest routes while also remaining efficient. The systems it has trained take into account weather predictions, driver feedback, historical data and more

And by using generative AI, the spokesperson said, drivers can more quickly navigate large office parks, apartment complexes and confusing terrains, all based on what the systems know and can comprehend about the locations in question.

“In many countries, such as India and Japan, street addresses don’t follow a standardized format. Some locations are described informally: ‘Third house on the right after the barn’ for example. With generative AI, we can derive precise map locations from non-standard addresses. That makes it easier and faster for drivers to find specific locations,” they said.

Walmart: Review summarization and product comparisons

Jon Alferness, Walmart’s chief product officer, shared in May on his LinkedIn account that Walmart is using generative AI for product and review summaries.

Rather than reading about all of a product’s specs or combing through dozens of reviews, customers
can use the AI-powered summaries as a starting point.

“These new tools capture the essence of customer sentiment for the most spoken of product aspects within the thousands of textual reviews allowing customers to quickly understand the overall sentiment and make more informed decisions,” Alferness wrote.

If the customer has a question the summarization doesn’t answer, they can dive deeper into any attribute of the product.

Alferness announced on LinkedIn in April that the company has also started using AI in its app to allow customers to compare up to four similar products by attributes, price, ratings, fulfillment options and more.

Walmart: Generative AI for search

The Bentonville big dog announced earlier this year it would launch a generative AI-powered search tool to enable its customers to quickly find and order products on its iOS app.

Traditional search requires a user to enter a specific product or brand, but Walmart’s generative AI search feature allows users to ask questions, like, “Help me plan a disco-themed birthday party,”
“Show me what to wear to a wedding,” or “What do I need to host a Super Bowl party at my apartment?”

According to the company, the tool can also take into account location, search history and other context about each user to provide personalized results for customers using the feature.The company said it expects the feature to offer in-app shoppers a more efficient, convenient experience.

Walmart launched its new search function using its proprietary data paired with LLMs and technology from Microsoft. It has plans to extend the generative AI-enabled search to Android and its website later in the year.

According to a June press release, it is also beta testing a conversational chatbot for stronger product discovery.

Walmart: Inventory management

Walmart said it uses ML- and AI-powered inventory management to ensure products make it to the right place, at the right time, in the right quantities. It made particular note of the technology during the 2023 holiday season.

The company uses its historical sales data, as well as engagement metrics like e-commerce searches and page views, paired with third-party data like climate and weather patterns, local trends, demographics of specific locations and more. Once those inputs go into the model, it can help predict the type of demand that will be associated with specific products, both during holiday season and throughout the year.

“By the time our customers are ready to shop, our AI/ML data has already completed the heavy lifting to improve inventory flow,” Parvez Musani, senior vice president of end-to-end fulfillment wrote in a December blog post.

Walmart: Warehouse automation

In April, Walmart announced it would roll out 19 autonomous forklifts in four of its high-tech distribution centers in partnership with Fox Robotics.

The forklifts unload pallets, moving them into an area where another automated system can catalogue and sort what’s on the pallets. The company said the human DC employees help determine the most optimized way to unload the pallets, even though they no longer have to do so manually.

“I’m watching players become coaches, and I couldn’t be more impressed,” Maurice Gray, a general manager of the distribution center where the FoxBot forklifts were first tested, wrote in a Walmart blog post.

Walmart: Middle-mile optimization

Walmart released its proprietary, AI- powered Route Optimization software to other businesses in April, with the goal of simplifying middle-mile supply chain issues.

The technology takes into account traffic conditions, customer locations, delivery times and more to show drivers the most efficient routes for pickup and delivery. The company said that, in doing so, it also decreases its carbon footprint. As of earlier this year, it eliminated about 30 million unnecessary miles, reducing its footprint by about 94 million pounds of carbon dioxide.

Emily Schmid, channels and platforms senior director, digital strategy at Walmart, said Route Optimization has helped place products in consumers’ direct paths, in turn cutting down supply chain inefficiencies.

“With its ability to organize and manage loading smartly, Route Optimization has made work more efficient for our associates,” Schmid told Sourcing Journal earlier this year. “It has improved the experience for our drivers by removing unnecessary wait times during their journeys…Our core purpose is to save people money so they can live better; today, that not only extends to our customers, but to other businesses and to the communities that we serve.”