In this exclusive interview, Sunzay Passari, Senior Director & Head of Innovation & Transformation at UPS, shares practical examples of how the logistics giant leverages AI and generative AI technology to boost growth, efficiency, and customer experience. He also offers valuable advice on determining the right technologies to invest in, challenges with AI adoption at a legacy organization, and what the future of AI holds.

Sunzay’s Key Takeaways

• “It always goes back to what business value we deliver to the organization. That becomes a key driver in determining what AI initiatives we invest our resources into. Business value is not necessarily just profit or revenue. It’s a combination of different metrics.”
• “AI is exactly where it naturally should be. Every new technology goes through a cycle of high boom and burst, and there is a lot of learning. As the end users and developers of the technology, we have our learnings that keep getting refined every minute.”
• “The most important thing, and what has been successful for me, is discovering and defining the real use case within your organization because many people ignore it.”

Want more insights? Watch the virtual insights session, AI-Powered Logistics: Path to Tech-Driven Efficiency and Excellence.

Hi Sunzay! Tell us about your role.

I’m the Senior Director of Innovation and Transformation at UPS. To understand the role, let me give you a little context. UPS is a 117-year-old company that exists in 200 countries. We own around 300 large jumbo jets, three airports, 150,000 trucks, and have 650,000 employees worldwide. This large scale brings many challenges and opportunities to drive innovation. Currently, I’m driving innovation and transformation initiatives for UPS. For the past 25 years, I’ve been driving transformation innovation initiatives across multiple domains globally and doing exciting things.

In UPS, what role has AI and generative AI (genAI) played in the past few years?

UPS has been one of the leaders in deploying AI solutions. One of our famous use cases is dynamic route optimization. We have achieved huge efficiencies in risk management for our employees and drivers. UPS has been a leader in route planning and weather forecasting using AI. If there is an urgency for delivery, how do we minimize our transit times and in the process, also reduce our fuel consumption?

But we’re not stopping there. With the emergence of AI and genAI, we are doing more from the logistics space to predictive analytics for demand forecasting.  There are technologies for automated sorting and movement of packages in the warehouses, and inventory planning inside the warehouses. That’s on the operations side, but an equal amount of technology is infused into customer experience as well, where there are many generative opportunities.

How do you determine which AI technologies to invest in?

There are always opportunities to infuse AI and incorporate AI. It’s important for us to be cognizant of what value we are trying to deliver, and we are a publicly related company, so we have a more conservative approach.

It always goes back to what business value we deliver to the organization. That becomes a key driver in determining what AI initiatives we invest our resources into. Business value is not necessarily just profit or revenue. It’s a combination of different metrics. For example, Thanksgiving is our peak season. All our resources are geared towards Thanksgiving, and we aim to bring smiles back to our customers. When a customer orders a package for themselves or their loved ones, we want to make sure it arrives on time without any damage.

Therefore, our business value is derived from customer and employee satisfaction. We have a huge fleet of people working at UPS, so we have our methodology for defining business value and what makes sense. While we prioritize AI, we are also experimenting with many things.

Can you give a practical example of using AI during hectic periods?

As I mentioned, the most important one is dynamic route optimization. Let’s look at optimization from an operational perspective. How do I increase delivery density in a given geography? How do I increase density in every truck moving from point A to point B? How do I make sorting in the warehouses more automated and precise? How do I give visibility to the end users on where their package is? While we are using many technologies, there is always human intervention in cases of unexpected incidents. People at every level you know, including myself and my boss, are all available. We make ourselves available to drive trucks, carry packages, and deliver packages when needed

Where is AI on the hype cycle?

AI is exactly where it naturally should be. Every new technology goes through a cycle of high boom and burst, and there is a lot of learning. As the end users and developers of the technology, we have our learnings that keep getting refined every minute. Today, the pace is tremendous. There is so much data and real-time insights available, so non-valuable propositions will eventually go out of the market, and all the valuable propositions will continue to grow. Of course, it’s overwhelming for us. Every company in the technology space has got a solution.

Where does this mindset come from?

Am I truly conservative? Probably not. While in my role of innovation, I can’t afford to be conservative, because when I got hired, I was expected to challenge the status quo. UPS is a large company, and we deliver over 25 million packages from Point A to Point B around the globe every day. With such huge operations, we must make sure that whatever we are doing, from a technology perspective, doesn’t create any disruptions. From a people and process perspective, with every infusion of technology, there is a lot of training and education required for the workforce to use it effectively.

The idea is that we don’t have to rush into things because our primary job is to bring smiles back to our customers. Sometimes we don’t do it for a business value, but for brand value instead. Maybe we are losing money, but it helps people perceive us as a cutting-edge technology company. Like any other organization, we have allocated resources and budgets. You don’t need unlimited resources; you must ensure that resources are allocated to the right things.

How do qualitative and quantitative research guide AI development to meet business goals?

It’s very important to understand our customer behavior. When I say customer behavior, there are two levels of customers. Let’s say, if you’re buying something from Amazon, that package gets delivered by UPS. So, you are a customer, but Amazon is a customer too. Therefore, we are deploying a lot of AI to derive insights for customers. Secondly, dynamic pricing. We must ensure we do very subjective pricing today with the technology.

Another AI-driven deployment on the customer interaction side of things is a customer service bot. For example, customers can interact with us to find out why their package is delayed. So, there is natural and precise interaction happening with the customer. Customers mostly interact with a bot, and we have infused a lot of data into it. However, we make sure there’s a human in the loop if the bot fails to answer any queries.

How does UPS balance customer satisfaction and data privacy?

For the last century, UPS pretty much knows every personally identifiable information (PIII) in the US (name, home address, phone number). But have you ever seen any legal case filed against us for either leaking or misusing any PII? We treat customer information with utmost care and security. We are very sensitive to our data. As we exist in 200 countries, we have to comply with potentially 200 different kinds of regulations and laws.

The World Economic Forum indicates AI can help reduce the carbon footprint of supply chains and logistics. Are there any practical cases of this in UPS?

UPS is focusing on two things. Firstly, our route optimization is not only about faster delivery but also fuel efficiency. Fuel efficiency is not only about saving costs but reducing negative environmental impact. These are part of our sustainability goals which we are investing a lot of money.

Secondly, we are one of the pioneers of bringing electric vehicles and drones for last-mile delivery, especially in Europe. We are pretty cognizant not only of the core logistics and operations sides. Let’s look at facility management. Our warehouses consume high amounts of energy as there is a lot of automation. Behind all of that, plenty is happening in terms of computing, servers, and infrastructure. So, in every chain of our process, we make sure that we are doing the right things; so that we can contribute back to the environment and drive the sustainability part of it.

With AI, we are getting more precise with our demand forecasting, inventory planning, and fleet planning. Also, we work with our customers very closely. By analyzing both customer and business data, we can decide the best steps for demand forecasting, inventory planning, or route planning.

Have you seen employee resistance to AI?

Yes, we have seen challenges but there’s more excitement and opportunities. As an organization, we make sure our training processes and content are strong. Logistics are pretty complex. While moving a package from Point A to Point B sounds straightforward, what happens behind the scenes is complex. So, we invest a lot in training. There is apprehension and resistance at times. Sometimes people get used to what they are doing for a long time. This is something that I have seen within my team and other teams as well.

For example, our route optimization is happening on a device similar to a small mobile phone. All our drivers have this device. In the back end, we have infused a lot of intelligence to deliver what is seen on the front end. Often, a lot of this technology infusion has been seamless from an end-user perspective but might not be the same for other teams. For example, our pricing analytics team took a few months to get used to new products as they were comfortable with existing processes and tools.

What is your advice for progress in AI innovation?

I encourage business leaders to keep themselves abreast with the advancements in the technology. It’s happening so rapidly that it becomes overwhelming, but at the same time not intimidated by what is happening. The most important thing, and what has been successful for me, is discovering and defining the real use case within your organization because many people ignore it.

What are the key AI developments in the next 5 years?

I’m looking forward to artificial general intelligence (AGI). So, I’m curious about how it will leap from the existing AI and genAI domain to both general kinds of intelligence, and how it contributes back to the whole planet, not just us.  Secondly, 5G technologies. 5G is not just about a network or a higher bandwidth. There’s a lot more than that. 5G is almost like a new dimension, as I keep talking about in terms of network and connectivity technologies.

*The interview answers have been edited for length and clarity.