ESI Interviews

Ep 36: Leveraging AI to Transform the Future of Global Logistics with FedEx EVP & CIO Rob Carter

Guest Michael Keithley
Rob Carter
March 6, 2024
Listen to this episode on your favorite platform
Apple Podcast Icon - Radio Webflow TemplateSpotify Icon- Radio Webflow TemplateGoogle Podcast Icon - Radio Webflow TemplateApple Podcast Icon - Radio Webflow TemplateApple Podcast Icon - Radio Webflow Template
Ep 36: Leveraging AI to Transform the Future of Global Logistics with FedEx EVP & CIO Rob Carter
ESI Interviews
March 6, 2024

Ep 36: Leveraging AI to Transform the Future of Global Logistics with FedEx EVP & CIO Rob Carter

On the 36th episode of Enterprise Software Innovators, Rob Carter, EVP & CIO of FedEx, joins the show to share fascinating insights into how FedEx uses AI, the future of automation in the transportation and logistics industry, and lessons on building a culture of innovation from his 23-year tenure.

On the 36th episode of Enterprise Software Innovators, hosts Evan Reiser (Abnormal Security) and Saam Motamedi (Greylock Partners) talk with Rob Carter, CIO of FedEx. FedEx is a multinational courier delivery services company with over $90 billion in annual revenue and is the fifth largest employer in the US, with more than 500,000 workers globally. With billions of packages delivered every year, using the world’s largest cargo fleet, FedEx uses sophisticated technology to provide a unique delivery service to customers across the globe. In this conversation, Rob shares fascinating insights into how FedEx uses AI, the future of automation in the transportation and logistics industry, and lessons on building a culture of innovation from his 23-year tenure. 

Quick hits from Rob:

On what the potential of AI means to Fedex: “We're at a fascinating point in the journey of AI where it's not just a buzzword but a tangible tool that's reshaping how we think about movement, connectivity, and the global economy. At FedEx, we're harnessing this potential to redefine the delivery experience.”

On the importance of understanding data: “The information about the package is as important as the package itself. This was a visionary statement by Fred Smith that has guided us for decades, and it's never been more relevant than now, in the age of AI and big data, where information is indeed power.”

On the importance of culture for innovation: “Building a culture of innovation isn't just about introducing new technologies; it's about creating an environment where questioning the status quo, exploring new possibilities, and taking calculated risks are part of everyone's job description. That's the culture we cultivate at FedEx."

Recent Book Recommendation: How to Win Friends & Influence People by Dale Carnegie

Episode Transcript

Evan: Hi there, and welcome to Enterprise Software Innovators, a show where top tech executives share how they innovate at scale. In each episode, enterprise CIOs share how they've applied exciting new technologies, and what they've learned along the way. I'm Evan Reiser, the CEO and founder of Abnormal Security.

Saam: I'm Saam Motamedi, a general partner at Greylock Partners.

 Evan: Today on the show, we’re bringing you a conversation with Rob Carter, CIO of FedEx.

FedEx is a multinational courier delivery services company with over $90 billion in annual revenue, and is the fifth largest employer in the US, with over 500,000 employees globally. 

With billions of packages delivered every year, using the world’s largest cargo fleet, FedEx uses sophisticated technology to provide a unique delivery service to customers across the globe.

In this conversation, Rob shares fascinating insights into how FedEx uses AI, the future of automation in the transportation and logistics industry, and lessons on building a culture of innovation from his 23 year tenure. 

Well, Rob, uh, first of all, thank you so much for joining. Uh, really excited to see you and, um, you know, Saam and I were really looking forward to this episode. Maybe to start, do you mind giving our audience a bit of background about your career and, uh, your current role at FedEx today? 

Rob: Yeah, happy to Evan, and it's great to be here with you guys.

I've been at FedEx for 30 years now, and I've been the CIO for almost 25 years. So kind of a land speed record. There's not a lot of enterprise CIOs that have been in the role for that long. I was the CTO prior to, uh, taking the CIO spot. Um, but, uh, really a cool arc of the company, huge amount of growth.

The world has become smaller and more connected across those 30 years and, uh, the technology has been a huge part of that growth and kind of the excitement of connecting the world and people and possibilities. 

Evan: I doubt any listener has not received a FedEx package in their lives, right? So I'm sure they're somewhat familiar with the organization. But do you mind sharing a little bit more about just some of the scale and complexity? Because I'm sure there's a lot of work you and the team do that, you know, is not fully appreciated from the outside. 

Rob: Yeah, from a physical network standpoint, you know, we have more than 700 aircraft. We connect 98 percent of the world's GDP and, you know, a day or two of connectivity there.

We have more than 200, 000 surface vehicles on, on the planet. And, uh, working hard to make them as sustainable and efficient as possible. But when you think of the way that we connect the world and move approximately 15 million packages a day on a typical day, it's really like carpooling at the end of the day.

You know, it's very efficient to load everything into a single vehicle and make dense routes and then stops happen. It really cuts down on traffic when you think about it that way. But It's a very cool company that people do count on. And it's one of the things that makes it fun to talk about is that everybody does have some experience with it.

Saam: You know, it's hard to have a conversation in December 2023 without at least talking about the possibility of AI is that next wave. And so maybe it'd be great to spend a few minutes talking about AI and and I'd love to start high level just to understand your take on where are we in the AI journey.

And, you know, if you think about the classic kind of hype cycle metaphor. Where do you think we are? 

Rob: Well, general artificial intelligence is, you know, is pretty new and it's coming at us really rapidly. But you know, we've all been dealing with versions of AI. You know, I can certainly remember, you know, writing some rule based AI back when, when I was a coder.

But, you know, the, the work at DeepMinds, the work, you know, that's really changed general intelligence computing and, and conversational things like chat GPT is really what's causing all of that interest because it's becoming so approachable now. It's not something that sort of lives, you know, back in the dark halls of coders anymore. It's right in front of us and we can all tap into it in an interesting and relevant way. 

For us, you know, I would tell you that we are leaning in. There's no question about that as we optimize our Networks and ability to deliver for the world. I mean things like machine learning and AI are incredibly important for us to Optimize what we do make it more efficient, you know, make the service better all of those things. But a lot of what our team members are feeling is that AI is coming to them.

Saam: If you take kind of a multi year view from here, are there ways that you think you may begin to incorporate AI into the FedEx business that might surprise the average listener. 

Rob: Yeah, I think so much of the, you know, so much of the, even if you want to call it a hype cycle right now, is that human to machine interface that exists out there and some of the, you know, some of the really more interesting things are The way we operate our enterprise is with massive connectivity, massive data creation every single day.

We have a digital twin that emerges each and every day because every time we touch a shipment, every time an airplane flies, every time a truck rolls, every time we call on a customer or they call on us, we capture data about that. And so we feed that data into a massive data lake, and that's another, you know, really important place where we're using AI, is as we literally capture petabytes and petabytes of data, we can look out at the business through the lens of AI and ML. And it's not a human to machine interface at that point in time. It's just really big data sets that make large language possibilities capable because we have so many scenarios. 

If you're shipping something internationally and you need to know the right clearance documentation necessary for that particular shipment from one part of the world to another, the AIs can, can tell you that. They can optimize that so you don't get frustrated with things being stuck in customs or surprise, you know, duties and taxes at the end of that. We can do a much better job with AI of activating the capabilities that come from massive data sets with use case after use case, and a lot of repetition when you think about it at the end of the day with 15 million shipments a day, we've kind of seen it all. So, you know, we can summarize that uniquely back in ways that allow us to optimize the airline, optimize the surface networks, optimize the pickup and delivery routes.

As, as a vehicle comes into your neighborhood or into a business district, um, we can be very efficient because we've been able to look at what's coming and pre plan exactly what the, what the best operating sequence for that day would be. 

Evan: So Rob, I feel like everyone in the world for the last like year or so is talking more and more about generative AI and there's, you know, these kind of new human computer interfaces. But I imagine for, for FedEx, I mean, you guys have been using machine learning and, you know, kind of algorithmic optimization for like a very long time, how are you using some of, um, you know, kind of more, you know, uh, predictive AI and machine learning to optimize, you know, to further optimize your guys business.

Rob: Yeah, I want to even zoom out from that a little bit, Evan. It's a super good question, but you know, when you think about the evolution of technology, it started with giving us a really pretty good view of what happened. You know, it was a record keeping thing. And then it got better and better at telling us what was happening, you know, moving closer to real time.

And then from there, it began to actually predict what might happen next. The, the, that generation. And then, you know, and then we almost have this clairvoyance world happening now, where it actually knows to ask questions that you haven't asked yet. It's, it can be clairvoyant about saying, have you noticed that it would be, you know, interesting to go look at this space because the data is telling us that there's a better way to run the network. So I love thinking of the evolution of technology across those different classes of how we handle data. And they really describe exactly what you were talking about. 

The traveling salesman, you know, linear algebra world that, that I first grew up and was about, you know, matrix solutions, and it was hard to get matrix solutions that were big enough to solve for the kinds and numbers of routes and networks that we were operating. Today, not so much. 

You know this. I'm not sure if this is really the era of big data or if this is the era of big compute. You know, we've had a lot of data for a long time, but big compute allows us to solve things that seemed unsolvable in the past. You know, really looking out at massive data sets and then really pinpointing things that predictively, you know, make sense for us to operate differently tomorrow, or to give our customers like what you may have experienced, you know, much, much better estimation about when, you know, when the package will be there and, and give you a lot more visibility about what's going on with your discrete shipment, um, through a network that's moving millions of shipments at the same time.

Evan: Rob, you talked about, um, You know, these new technology, or sorry, I guess kind of maybe older technologies, but now unleashed through the, this kind of age of compute, being able to solve problems that were kind of unsolvable in the, in the past. Can you help us like dream, help us dream a little bit about the future, right?

Are, are there kind of upcoming use cases you see as technologies being applied to that you feel kind of bullish about? Um, even if maybe some of your peers feel like it, so maybe in the science fiction realm. 

Rob: Well, there, there are a lot of things. I mean, I love autonomy. I think autonomous vehicles are a big part of the future, no question about that.

And, you know, most of us tend to think about, you know, Waymo and around city driving. But one of the most important forms of autonomy that's coming is over the road driving. Those are difficult jobs. They're, You know, and, and human in the loop is probably still really important, but, you know, having all the aids, many of us have them in our cars now, you know, whether it's adaptive cruise control, auto braking, you know, a lot of, a lot of the things that allow for autonomous driving that make us all better drivers, frankly, not less attentive, but, you know, but better because all the machine vision and computer vision around us, the, the, the LIDAR, all of those things, they, they don't get, you know, they don't get thwarted by being tired. They don't get thwarted by fog and, and rain and, you know, difficult driving conditions. And so that, and the fact that, you know, we can see a lot more about the road than we ever could.

I mean, as, as all of us go around with Google Maps, or ways or whatever, you know, it's like there are thousands of sensors out in front of us that are, you know, telling us before we see the brake lights up in front of us and go, uh oh, there's a problem that there's a problem. So, you know, the connected world, the IOT, you know, maybe the most. common way to talk about it is, is still really cool to me, but you have to even take that and extrapolate it all the way down to the individual package. I mean, all of these shipments from all the places that you know are going to become a sensor based shipments. They're going to have some form of RFID or Bluetooth, low energy tags on them that will allow them to be on the network all the time.

So you'll be able to, Continuously, not just see the last time your package was touched, but where actually is it and having that kind of visibility. We did that, for example, on the vaccine shipments during COVID. We have a highly patented technology. We've got about 80 or 90 patents on our sensorware technology, which is a Bluetooth low energy.

Very small, you know, call it a little box of chicklets, little two chicklets or something like that, that you put on a shipment and it connects to all of our network. It connects to our handhelds, all of our wireless access points, all of those things bring those shipments on. So even when operations were extremely difficult, uh, during COVID. We were hitting 99 plus percent reliability for vaccine deliveries all over the world because we could see it.

We could see if something had gone wrong. We could even see if it got left behind and take remediation and intervention actions on it. So, you know, that kind of connected world, I think, will Um, you know, it'll come, you know, it'll probably seem normal to us because we're so used to the, the internet of things these days.

But you know, when, when it becomes the, the rising law of the, the law of rising expectations, it'll be an important part of what we all expect in a connected world. 

Saam: One thing you touched on earlier was the importance of data. And as we all know, these AI applications get delivered on the tops of of data and data sets.

And I think there are a lot of technology leaders who we bring on to the show or who are listening to the show that may be leading IT and technology teams at large enterprises that have the benefit of significant data sets that are relevant to, you know, their products and their customers, but that are asking themselves the question, how do I unlock the power of this data in AI?

And then they realized there's a lot of infrastructure investment they need to go make. And so, you know, for the listeners who may be sitting in that position, maybe a few years behind where FedEx is on its AI journey, are there two or three kind of concrete recommendations you would make around how to develop a data and infrastructure strategy to set yourself up to recognize and realize the potential of AI?

Rob: Yeah, another really good question. I believe that, you know, sometimes the mistake we made, we make is that believing that our systems of interaction, systems of operations, systems of, systems of engagement are the places where all that analytics takes place. And the reality is, is we pump massive amounts of data from those same systems of interaction and operation and engagement into a data lake.

And so once, once you take the data and Manage it in the different constructs, it can be much less structured. You know, it's you still don't want it to be a data swamp, but you don't want it to be a data puddle either. Put, you know, put the data out there and then let these next generation compute and tools build machine learning algorithms that refine your view of how your operations are working.

Another thing you can do once you get it out there in, in those places, and we just happen to use Azure and, you know, a lot of Power BI around that and, and the capabilities that are there, but, you know, they all, they all have great, great tooling, but you can also then bring in, uh, external sources of data.

So we bring in traffic data and weather data and geopolitical things that, you know, can affect our network. There's so many things that you can then put in. to a very large data set like that. And, you know, I, I cut my teeth on really structured data, normalizing data, managing data to, you know, to the finest level of granularity.

And what I've had to unlearn, you know, about the modern world is that you don't actually have to do all of that. You can, the data kind of resolves itself when you think of even web search right now. You know, nobody's out there normalizing their websites, but the, but the search engines can take that massive amount of information and make it make sense to you, you know, with a simple query.

Same thing in AI, same thing in, you know, in power analytics out there now is that you can, you can have the data be a little bit messier. And so a lot of times when I'm talking to my colleagues out in the industry and they're asking that same question, I tell them, look, go a little quicker, you know, be less. Be less hypersensitive about, you know, exactly how to manage the data. It doesn't mean you, you know, you don't have, you have to protect it, by the way.

I mean, we do that well. It's our data. It's not going out into public spaces, and cloud does not mean secure automatically. So we apply, you know, information security and, you know, all the things that are necessary to protect PII, all of those things out there. But the data itself has a way of self organizing much better now than it ever did. So you can go fairly quickly, but don't, don't just attach that to your IT systems because it's a different beast. 

Evan: Do you have any kind of advice for our listeners around how you kind of build a culture of innovation? How do you get your team to embrace new technologies while realize we'll kind of like appreciating the realities of the world you live in?

Like any advice about how you kind of lead the team from a cultural perspective? 

Rob: Yeah, Evan, we're, we're really fortunate in that innovation is in our DNA. Fred Smith, who founded the company and, and ran the company up until last year as the founder, chairman and CEO. He's executive chairman now at the company.

He said in 1978, he said the information about the package is as important as the package itself. Well, what makes that such a, you know, kind of a mind blowing statement in 78 is that, you know, people were going like, what information? What, you know, I mean, because transportation systems were essentially black holes that things went into and popped out at the other end.

But, He was the visionary that said, no, we, we should be able to use technology to see each step of the progress. And he redefined what inventory was at that time. Inventory turned into not just product at rest, but product in motion and at rest. So, you know, in many ways, the inventor of just in time manufacturing systems because our big commercial customers could predict with a high level of certainty exactly when critical parts would arrive and so they didn't have to stack them up. Well, that DNA worked its way into, into our company.

When You know, in, in the late 1990s, when, when we were first standing up, um, FedEx. Com, I mean, you know, every, every website that was in that earliest part of, you know, of enterprise websites was an about page with a message from the CEO and, you know, our very first page was a package tracking page. It was an HTML page that had a blank in it. A little, a little gif with a flying package going across it said, enter your tracking number here.

And so we actually won a Smithsonian award for that, for being the first business transactional website. Because you could actually do something with it. Well, we were able to do that because we were already doing it on PCs and with, you know, with EDI, it would be a little bit of a stretch to call them APIs at the time, but our customers could connect to us and pull tracking data, you know, to, to their needs and, and put it to work.

And so through that, you know, whether it was the internet boom or the cloud boom, or even, you know, before that, the distributed computing boom, where things weren't just all locked up on the mainframes anymore, we developed this culture of utilizing technology for strategic advantage. Utilizing it to operate more efficiently. Utilizing it to grow by giving customers things that, you know, became sticky for them and that they, they enjoyed using.

So I feel really good about the nature of innovation that Fred built into the company and that many of us have been able to carry forward over these years. 

Evan: And Robby, maybe one more question, kind of more organizationally, like, um, you know, as technology is, you know, technology is obviously changing very quickly, right?

And I have to imagine there's probably even more use of data, machine learning, AI, in FedEx's future rather than its past, what are maybe the implications and how you think about developing the organization? Are there new capabilities or new training or new tools? Like, how do you, yeah, how are you kind of planning ahead to help, you know, help make sure your team is ready for that unknown technology future?

Rob: Yeah, you know, it's, when you have a long legacy of technology, it might either be your best friend or your worst enemy, you know, because old habits die hard. And you know, so we're, we're at the very end of the mainframe era for FedEx. We're down to, you know, we've, we've limited 98 percent of the MIPS that, that we once ran on the mainframes.

We've got a couple of, you know, kind of obtuse applications that still run out there. But we're, you know, we're, we're having to break those habits really hard because it's not the dominant design of technology anymore. You know, you can tell what dominant design is by looking at a Greenfield company out on the West coast where, where you guys are and say, what are the odds that a startup company today would build a data center, much less buy a mainframe, and the odds are exactly zero. That's not the dominant design anymore. So we're exiting, we'll be out of all of our data centers worldwide next year. Um, nothing left in, in the data centers. We'll have some own technology and exascale colos that are out there, and, and, you know, a great deal of it in the cloud as well.

But we're, we're actually having to pry ourselves forward, and the real answer to your question is what does full stack software engineering look like these days? How do you use DevSecOps and CICD and the capabilities of modern software delivery to generate speed to value? Because the hardest thing for enterprises and in a lot of this audience of yours or enterprises are speed to value.

The complexity and sort of tangle of the past waves of technology, that have crashed on the beach over decades, may still be out there and it may be holding you back because you've got processes around those. You've got data locked up in them, so we're heavy lift. Probably the hardest thing I've done across the arc of my career is constantly modernize and constantly convince the business that the investment in modernization and technology is worth it. Because, you know, businesses just want, you know, they want function, they want value, they don't really want I. T. for modernization sake or data for data sake. They want to, they want to see us moving the business forward. So striking the right balance between modernization work, which is really engineering work and delivering value for the enterprises is hard.

Evan: And Rob, um, as you guys are kind of like preparing for this future, help us dream a little bit. Like, what is, what is it like to be a, you know, FedEx customer or FedEx employee, you know, five years from now, 10 years from now? What are some of the, what are some of the things you think that, uh, you know, could be possibilities in terms of, you know, how are you using these technologies to either optimize how you run the business or to improve the customer experience?

Rob: Yeah, there's, there are a ton of fun things there. I mean, think frictionless though. Think that, you know, these things just really kind of happen for you. Instead of customer service and it's traditional, you know, somewhat frustrating format today that there's, you know, there are a lot of chat options available to you that use gen AI capabilities and natural language to, you know, give you quickly the answers that you need, or at least ride copilot with the, with the customer experience team so that they can give you the answers that you need more and more quickly and efficiently.

Think about sensors on the packages. I already mentioned that, but If this was a video podcast, I could, I could show you a sensorware and it's got a little loop. You could put it on your dog's collar, you know, and have it run around with that because we give you access to that Bluetooth low energy device and the year's worth of power that it still has left in it when it arrives at your doorstep. You know, think of, you know, delivery robots and autonomy and, you know, things that just, you know, make, you know, make it much greener because they'd all be electrified. We're, we're electrifying all of our pickup and delivery vehicles so that, you know, when, when we're in your neighborhood that, uh, we're leaving a clean, a clean footprint behind us.

You know, it's, it's a brave new future, but it's all about this connected world that we live in now, you know, it didn't used to be that way. Distribution systems used to be, you know, from manufacturer to distributor to warehouse to store, you know, into the car. Now they're much more discrete systems that go from the point of manufacturing directly into your hands.

You know, a lot of times if you order a piece of electronics, new iPhone or something like that. You can literally see its journey and it's, and it's so much more efficient that way. There is not a lot of intermediate distribution points or middlemen that want a piece of the pie. It's efficient, it's clean, and I think it's exciting.

I mean, we all learned a lot about it in COVID because, You know, whether we were waiting for, you know, critical hospital supplies or vaccines to show up, or whether we were waiting for dog food to show up on our doorstep, we, we, we all learned a lot more about the, you know, how cool it was to be able to click and get and, you know, that's, that's just, uh, you know, more of that to come in the future as we do, you know, more and more integrations with our, with our big customers and more and more capabilities for the recipient.

Evan: Well, Robert, at the end of the episode, we got probably like 10 minutes left. Um, like to have a quick lightning round of like, you know, four or five questions. And, um, the, these aren't the easiest questions to respond, to respond to concisely, but trying to look for like kind of the one tweet, the one tweet version.

Um, so, um, Saam, do you want to kick it off first? 

Saam: Absolutely. So Rob, to start, how do you think companies should measure the success of a CIO? 

Rob: Value delivery, period. You have to deliver value for the business and you have to measure that value.

Evan: Rob, you've been in your role for a long time. You've seen kind of, you know, such a unique perspective and experience.

What's maybe one piece of advice you wish someone told you when you first became CIO of FedEx? 

Rob: Realizing that business partnership is as important as technical skills. When the business and technology teams pull in the same direction, this job gets to be easy, but when they fight and point fingers, it's an insult to glaciers to say that things move at glacial speed.

Saam: Rob, I think connecting to Evan's point on just the length and success of your tenure at FedEx, I'm sure one of the things that you've refined is how you as a CIO interact and collaborate with the rest of the C suite and the rest of leadership. So what would your advice be to CIOs who are listening around how they should position themselves to work effectively with the rest of the leadership team?

Rob: Be a business person, not a technician. I've been lucky to be able to sit at that table now with giants like Fred Smith. I worked directly for him for 23 years. Be a business person first and help bridge the gap between what technology does, and what it doesn't do, and how to partner with the business to get the most out of it.

Evan: Rob, on a slightly more personal note, um, is there a book that you've read that's had a big influence on you? And if so, love to hear why. 

Rob: I loved How to Win Friends and Influence People and I've never been to a Dale Carnegie class in my life, but I love the vignettes in there that give you a little snippets of wisdom about how to situationally course correct when you need to.

Saam: Staying on the personal side, uh, what's an upcoming new technology, and it doesn't need to be AI related, that you're personally most excited about? 

Rob: I love electrification. I think, I think I'm a car guy, but it's so much fun to see what the next generation of mobility and transportation looks like. I like mobility and all the things that come with that.

Evan: Rob, final question. What do you think is going to be true about technology's future impact on the world that most people today would consider science fiction? Looking for your kind of contrarian view. 

Rob: Yeah, none of us could have ever predicted where we got to with technology. I mean, Nesbitt had it right when he talked about future shock.

My belief is we're going to be shocked by what comes next, and none of us are prescient enough to quite see it yet. 

Evan: Well, Rob, I really appreciate you taking time to join us today. I'm looking forward to chatting again soon. Thank you so much for joining us. 

Saam: This was awesome, Rob. Thank you. 

Rob: My pleasure.

Evan: That was Rob Carter, CIO of FedEx.

Saam: Thanks for listening to the Enterprise Software Innovators podcast. I’m Saam Motamedi, a general partner at Greylock Partners.

Evan: And I’m Evan Reiser, the CEO and founder of Abnormal Security. Please be sure to subscribe, so you never miss an episode. You can find more great lessons from technology leaders and other enterprise software experts at

Saam: This show is produced by Luke Reiser and Josh Meer. See you next time!