How Walmart’s CTO gets technology to adapt to customers

Mr. Kumar’s roles at the global retail giant include overseeing engineers at Walmart US, Sam’s Club, and Walmart International. Mr. Kumar is also Executive Vice President and Chief Development Officer of Walmart.

In an interview with Mint, Kumar, a graduate of the Indian Institute of Technology Madras, spoke about how Walmart is combining advances in technology such as artificial intelligence and generative AI with the company’s “adaptive retail” strengths. Edited excerpt:

Can you describe a specific example of how Walmart successfully implemented advanced solutions to improve operational efficiency and customer satisfaction?

Every aspect of the industry is undergoing dramatic change. Previously, people would go to (physical) stores to find products. Then, with e-commerce, you enter a few keywords and a list of products appears. Now, things like social commerce such as “Shop with Friends” are pushing the product discovery aspect far beyond traditional e-commerce search.

(Walmart’s “Shop with Friends” feature allows for virtual fitting rooms that can be shared across messaging and social platforms like Pinterest and Instagram.)

In India, payments and identity (pronounced Aadhaar) have a huge impact on commerce. Supply chains are also evolving dramatically. It’s not just about drones and self-driving cars, it’s also about how the products themselves get from manufacturing to customers’ homes and even their refrigerators.

Walmart uses technologies ranging from augmented reality, robotics, digital twins, AI, and now GenAI. How do you connect the dots for your customers?

What ties all of this together is adaptive retail. It’s about how technology adapts to customer needs. With social commerce and his GenAI as an example, it’s important to reduce customer friction as part of the shopping mission. Most of what we’re trying to do with technology is to remove friction from the customer experience of finding, transacting, and receiving products.

What processes are you using to leverage analytics and AI to improve the customer experience?

Our processes are truly customer-centric. That’s how we always start our meetings. We don’t start by saying, “We have some nice, cool technology. Let’s figure out if it helps our customers.” Instead, we look at our customers’ usage patterns and speak directly with them to gain information about their likes and dislikes.

We also have vast amounts of data on our customers, their online shopping habits and preferences, when they enter a store or pick up a delivery. We collect all this data (from multiple channels) and utilize several AI models built in-house, as well as the best available on the market (Large-Scale Language Models, or LLMs) .

We can train (these LLMs) on data, build our own custom models on top of it, and glean insights from it. Based on these insights, we start designing experiences based on what we think are the most important problems that need to be solved for our customers.

Essentially what we want to do is continue to close the loop between what the customer wants and what is the best technology to solve that problem, implement the technology, and go back and do it. See if it was helpful and keep improving it.

Can you give me an example of how this process has helped your customers?

Consider the example of home delivery. This started as a relatively small project or proof of concept. When a customer places an order, we send a representative to deliver the product not only to their home, but also within their home. And in your fridge too. This requires a lot of trust from the customer.

You also need to understand the replacement settings. Let’s say you order orange juice, but due to a surge in demand, it may be out of stock and you need a replacement. So we listen, and we go back and solve the problem. This is an ongoing process and helps improve the customer experience.

How is GenAI helping Walmart improve conversation productivity, efficiency, and effectiveness?

GenAI focuses on four broad areas: One is about customer experience. The second area is for many internal applications. We recently launched a product to help our in-store employees. It’s called a sidekick. This allows store associates to focus on the next best thing to do.

Stores are highly complex environments. People are running around and need to perform many tasks. When combined with internal models, GenAI can not only identify the best employee for a particular task, but also assign the task to that employee. AI and GenAI’s ability to understand context is helping improve employee effectiveness.

The same thing is happening on the corporate side. If you look at our supply chain, we have built a model that allows us to understand demand and optimize inventory flow. Our supply chain is very dynamic. When demand spikes in one region, we can now automatically find the best way to direct inventory to that region.

The third area is on the creative side, whether it’s creating personalized advertising and marketing campaigns for customers so that GenAI can understand their needs, or gaining deeper insight into their product catalogue.

Fourth, we use GenAI to build unique tools to improve testing, improve operations, and understand how to scale up and down systems.

However, when using GenAI tools, you also need to be careful about data privacy, security, illusions, etc.

yes. One of the fundamental pillars is trust. It’s important to make sure the technology we deploy is built on trust from our customers and employees.

We do not pluck LLMs from the wild and release them to our customers or employees. Because we also operate one of the world’s largest pharmacies, we have built an internal platform that creates data privacy and security safeguards to protect customer data and other sensitive data.

They also don’t want to compromise their supply chain data.

We then built an internal layer that helps you choose the best LLM for the task you are trying to perform. For example, we’re currently rolling out a pilot to help legal departments sift through the complex regulations they have to deal with. In this case, a generic LLM will not help. This can also help deal with hallucinations.

What role are Indian centers playing in the adoption of emerging technologies like AI and GenAI?

India is a great place for engineering talent of various natures, from software engineers to data scientists to product and UX (user experience). We have increased our presence in Bangalore and opened another location in Chennai during the pandemic. Our Indian base is one of the largest engineering groups in the world. However, we are working as one global virtual team.

Let’s take GenAI as an example. We have a very strong team in our IDC (India Development Center) that works very closely with our teams based in Hoboken, New Jersey and Sunnyvale, California. Their sole focus is to use LLM to produce the best search experience.

Are there synergies between you and other businesses like Flipkart, Myntra and PhonePe that share technology expertise across sectors?

Our subsidiaries in India are largely self-sufficient. But because we’re the majority owner (of these departments), we’re continually looking for places where it makes sense to share information, best practices, etc.

As an example, the team at Flipkart is working very closely with IDC and the US team to share learnings in areas such as GenAI.

What about the role of automation and robotics in improving productivity and customer experience?

We announced a major investment to automate large parts of our supply chain. This is important as we plan to build the future of supply chains and, more broadly, the future of retail.

The focus is on using automation, whether in the form of robotics or our automated storage and retrieval systems, to tackle difficult, repetitive tasks and do what employees do best: help customers with a human touch. services, allowing you to focus on serving your customers. Cases requiring human intervention.

We recently opened a cold chain automated storage and retrieval facility, ensuring that the entire facility is cooled to several degrees below freezing. This is not an ideal situation for most employees. Imagine having to carry large product cases while wearing unwieldy winter clothing.

Automation not only helps a lot in these cases, it can also significantly improve customer service and reduce the time it takes to get products to customers.

How do you address the sensitive issue of job losses due to automation?

We found that our employees were excited that we invested in automation because they saw us doing two things. One is to create a better working environment for them because the most dangerous tasks they hate the most are no longer necessary. And secondly, it gave me an avenue to improve my skills.

Therefore, they can make a career out of their work. Overall, I agree that the nature of work will change (through automation, AI, GenAI). But most of the time, things seem to change for the better.

Given the incredible speed with which technology is advancing, what advice would you give CXOs when implementing these technologies?

Technology evolves very quickly. And one of the challenges is keeping up with all these changes. Therefore, you need to keep three things in mind.

The first is securing and empowering a team of skilled and talented people.

Second, it is important not to get carried away with new breakthroughs in technology for technology’s sake. It can be distracting, so keep yourself grounded. The way we do this is to start with the customer, see what the real problem is, and then explore how technology can benefit them.

Third, start small. At Walmart, we do a lot of test-and-learn proofs of concept. For example, just because we had access to LLM, we didn’t switch our entire search algorithm.

Trade-offs and risks, such as regulatory issues and illusions, need to be understood. As the technology matures, it can be deployed more widely.