Driven by Data:
Transforming Transportation with Qlik and Penske

Leveraging AI and Data in Transportation

Michael Krigsman: Imagine a fleet of 400,000 trucks that predicts failures before they happen. Last year, Penske's AI prevented 95,000 roadside breakdowns and guided 117,000 repairs, turning its vehicles into data centers on wheels.

I'm Michael Krigsman, and this is CXOTalk. 

Joining me are Sarvant Singh, vice president of data and emerging digital solutions at Penske, and Chris Powell, chief marketing officer at Qlik, the data platform powering Penske's analytics. 

Today, we'll unpack how 170 citizen developers, finance analysts, maintenance experts, and other business pros build production AI apps and the practical steps your organization can take to achieve similar results.

Sarvant Singh: Penske is a leading provider of transportation solutions specializing in rental, leasing, logistics, and supply chain management. With more than 430,000 trucks and thousands of locations, we play a significant role in sustaining supply chain operations across North America.

So, in my role as the vice president of data and emerging digital, I focus on two primary things. One, making the right information available to the right people at the right time and, second, innovations and digital transformation through effective use of emerging technologies like artificial intelligence, IoT, cloud, virtual reality, and now generative and agentic.

Michael Krigsman: Chris, you're working very closely with Sarvant, and tell us about Qlik.

Chris Powell: Qlik's been around for 30 years now, so it's not a newcomer to the data space, and we're really in the data business. When we talk to customers, it usually falls into five areas. They have to be able to bring the data together from a lot of different kinds of data, different sources. They have to make sense of that data, transform it so it's usable. They have to make sure that it's the right quality, that they trust it, and then, of course, analyze it.

Analyzing it today ranges from traditional analysis all the way through much more sophisticated uses of AI and predictive analytics. And then the fifth is taking action with it and doing something, whether that's enabling a person to do something with it or fully automating the next step in a process. What's so fascinating about Penske is when you see their trucks driving down the road, you just don't realize how much data is driving all of this, creating a better experience for their customers and a better experience within their overall employee base. It's just a fascinating world that things have been evolving into for the last 30 years.

Michael Krigsman: Sarvant, talking about leveraging data to improve outcomes, you're a transportation services company. You mentioned you have over 400,000 trucks on the road. How are data and AI critical to your...?

Sarvant Singh: After our people, our vehicles are our most important assets. But they're not here in the corporate office; they're out there in the field transporting goods. And before IoT and AI, a problem could happen unannounced, without much warning. And vehicles these days are very complex. On average, they have multiple computers and hundreds of sensors. So, finding what is really going on with a truck can be very difficult for the drivers, and that's where Penske comes in.

We are collecting real-time sensor data from our trucks. We are bringing it into our data platform and then connecting it with our operational data and very rich maintenance history.

So, we implemented a predictive AI solution called Proactive Diagnostics, which can predict when a truck is going to fail.

We have been implementing AI in our business operations and back office functions for almost a decade. But over the last few years, we have started implementing AI in our front office functions as well. And we are launching customer-facing digital AI products.

Michael Krigsman: Can you talk a little more about those machine learning and AI models?

Sarvant Singh: Yes. Proactive Diagnostics, the predictive AI I was talking about, prevented 95,000 trucks from failing on the side of the road last year alone. So, that's the power of AI at work.

The other solution, Guided Repair, the prescriptive solution I mentioned, was used 117,000 times last year by our technicians to perform repairs. And apart from that, we have been implementing AI to reduce the administrative workload for our maintenance supervisors.

The most significant thing about this particular solution was that it was a self-service AI. We implemented a self-service AI capability almost three years ago, and that is helping our citizen data scientists build AI models to improve or transform business processes.

And this particular model I was talking about, which helped reduce the administrative work for our maintenance supervisors, came out through that self-service AI process. So, I'm extremely proud of that.

And then very recently, we implemented our GenAI platform. We have already implemented two GenAI products on top of that: one for improving service desk operations and a second for supporting the 24×7 roadside assistance call center. So, we are very, very proud of those implementations.

Building a Culture of Data and AI Trust

Chris Powell: We just had our QlikConnect conference in the last couple of weeks. And one of the things that came across over and over again was how many organizations have a strategy but aren't able to really implement that strategy. A lot of organizations have ideas, but then the analysts keep telling us how many of them keep failing. What's the secret?

Sarvant Singh: That's a great question, Chris, and I do get asked that a lot. We start by focusing on what's the right thing to do? What can deliver a great customer experience? So we really start from there and then we start small and then fail fast. Once we are successful, we want to scale very fast.

That's where building a robust data foundation and an ecosystem of partners is just so important because you can't do all of this alone. You really need an ecosystem of partners bringing different capabilities. And at Penske, we put all of that together to solve difficult, complex problems for our customers, and I think that's what this is all about.

The real differentiator for us is our culture. We have a culture where different stakeholders and teams come together, and we collectively work to first define the problem, then figure out the best way to solve it. And particularly when it comes to AI implementations, being able to combine our people capability with our data capability and emerging technology becomes the true differentiator. It is because of this combination that we could deliver all these solutions and, in the process, achieve competitive differentiation.

Michael Krigsman: Chris, you talk with many Qlik customers across a variety of industries. What are the broader trends in AI and data adoption that you see?

Chris Powell: The big thing is this shift from strategy to execution. So much of it is exactly what Sarvant said. The companies that are doing this really well understand that there is a lot of experimentation, but you need to do that fast failure, as Sarvant said, and move forward, scaling the things that work and discarding the ones that don't. AI is supposed to bring many efficiency gains, but organizations are not seeing those gains, so they need to find ways to actually realize them.

And probably the biggest challenge we hear keeping people up at night is that they just don't trust the environments they've been creating. So if you don't trust it, it doesn't matter what it costs or if you're getting efficiencies. You're not going to want to use it. Companies are getting closer and closer to understanding how they should implement it and where they should implement it. Now they're moving into this phase of asking, "Can they implement it?" Can they implement it given the cost? Can they implement it if they don't have the right people or enough people? Can they implement it in a way that's trusted? Those three factors, cost, efficiency, and trust, are the big trends that we see organizations trying to overcome.

Sarvant Singh: When it comes to trusting the data, we need to make sure that we are building the right data management practices. We are making sure there is data stewardship happening so that we can build this robust data foundation to support AI.

But when it comes to trusting the AI, it's a different ball game because there's a human element involved. That's why, whenever we implement an AI solution at Penske, we make sure to pull in our subject experts as our first line of defense. They become the people who are really testing and validating the AI. 

And so the way we position AI is not just artificial intelligence but augmented intelligence, and that has been the key. It's really about combining AI with human intelligence, and I think this combination helps generate the needed trust in our business processes.

Empowering Citizen Developers and Data Culture

Michael Krigsman: What is unique about Penske that enables this AI- and data-related business maturity that a lot of organizations just have not yet achieved?

Chris Powell: As Sarvant’s saying, they've been leading the way for quite some time, understanding the impact of data on their business. People don't think of Penske as a data business, but when I describe what Qlik is, I say we're a data business. I guess, Sarvant, in a lot of ways you could say you're a data business. It's almost as if the Penske organization is a data business masquerading in a different industry. 

And the foundation they've been able to set up over the years establishes an understanding within their organization of what data they need to bring together. The importance of that is probably recognized throughout their leadership team.

Sarvant, I'd be curious whether you're finding that this isn’t just an IT or business intelligence group that's thinking about this. It's not just the data people. It's the business that starts thinking about how data is going to transform our business, which requires that business partnership. If it's just the technical people, just the data people, thinking about this, then they don't have the partnership they need.

Sarvant Singh: One of the reasons we could achieve this success is because of the self-service analytics initiative launched years ago. The idea is to deliver data tools and capabilities and make them easily accessible to the business community. 

So we have a very distributed execution mechanism, and I think that is our superpower: we have our business analysts, data analysts, and citizen data scientists sitting in the business, where they're very close to the subject matter. And we have trained them to be Qlik developers, to be citizen data scientists. And when they are building these digital products, quality is given from day one because they have the needed subject matter expertise to make sure it works the right way and can be embedded as part of our business process.

Just to throw some numbers out there, we have roughly 170 Qlik developers and 50–60 citizen data scientists, and I very proudly say that these are Qlik developers and citizen data scientists I didn't have to hire.

Michael Krigsman: Sarvant, would it be accurate to say that your current success with data and AI relies on the fact that years ago Penske established a culture of data?

Sarvant Singh: You're absolutely right. That is the key ingredient to this success. As I mentioned, we have a group we call the Penske Analytics Network, and there are more than 300 members in it. Out of those 300, 170 are Qlik developers. And then we have this citizen data science community. What we have done is provide them with the right data, the right tools, and the know-how, and that has been the primary driver behind the groundswell we are seeing on the analytics side.

Chris Powell: So, Sarvant, your point about this data foundation, really this culture of data, is exactly what we see in the best-performing organizations. The self-service nature of this, and the number of people you mentioned who are developers and citizen data scientists, is a fascinating number.

Sarvant Singh: Absolutely, Chris. Collectively, these citizen developers have built more than 700 applications. Just last year alone, we saw almost four million logins. That's a very impressive number.

Michael Krigsman: When you say a "citizen developer," can you describe that person for us?

Building a Data-Driven Culture

Sarvant Singh: They may be in finance. They may be in marketing. They may be in maintenance. Their job is to support the needs of their own function, but what we have done is augment their skills in that function with technical skills. It may be skills on Qlik Sense. It may be using Qlik to analyze data in the data lake, or it may be skills to build AI models using automated analytics platforms. So it's a combination of those. 

Data storytelling is something we have been focusing on as an organization. We have also been focusing on things like data literacy and AI literacy, and I think it's the combination of these efforts that has influenced our culture and really helped transform it to become a data-driven organization.

Michael Krigsman: So these are folks who are working in various functions (like finance or others) who are subject matter experts in that area. These are not technology experts, correct?

Sarvant Singh: That is absolutely correct.

Chris Powell: Yeah, and that's what we see everywhere. The organizations that are able to really remove the bottleneck of what was traditionally a centralized business intelligence group. If you have a question, you go to that small group. If you need a change to an existing application, you have to go to that small group. It makes it very difficult to become a data-driven organization when you have that bottleneck in place.

So, having capabilities within your environment to enable these teams to operate self-sufficiently, but to do it within a trusted environment, is crucial. You don't want to have a wild, wild west out there. You need to have controls. You need to make sure you're still compliant. You need to make sure that the right people have access and that others do not.

Chris Powell: Having all of these pieces in place (the terminology often used now is "treating data as a product," or data products) means you can serve up these data elements in a secure way, in a way that, as the people responsible for data, you still have control. And it's not like the marketing department or others are now taking action on data you don't trust, which can have important consequences that you're trying to manage, avoid, and help people with. 

Empower people but make sure you have the right controls in place so things don't get out of control.

Sarvant Singh: We do have governance around self-service BI and self-service AI. We have a Data and Analytics Governance Council, and we're also putting together robust AI governance processes right now. We started that process last year and are making good progress. 

But, as Chris mentioned, making sure our citizen developers have the freedom to conduct proofs of concept and experiment with the data. We provide all that flexibility while ensuring it is in a safe and secure environment. We have the right guardrails and controls in place.

This becomes really important when we are operationalizing a solution, thinking about these digital product implementations in two phases. One is the proof-of-concept phase, where we have all the freedom we want. But then we have the operationalization phase, where we make sure that we put the right controls in place and the processes are hardened. Before something goes into production, it is truly enterprise-grade.

Leveraging Data and AI for Business Transformation

Michael Krigsman: Chris, how does Qlik's technology enable these kinds of initiatives and support that data culture which is so important to Penske?

Chris Powell: We're extremely proud to be a leader, and the only leader that cuts across data integration, data analytics, and data management. We're trying to hit as many of those marks as possible in order to make sure that customers can trust that data environment, bring it together, and serve it up for their organizations to get value from it.

Michael Krigsman: What strikes me, talking with both of you, is that it's not just the technology, but what happens when the technology meets a receptive group of people who can then enable the business outcomes.

Sarvant Singh: At the end of the day, even AI, like everything else, is a people business, and we have to get that right. I do want to highlight the importance of having the right people in place. That's why I'm very proud of the work my team has done and the work our citizen developers and citizen data scientists do on a daily basis.

Chris Powell: A number of years ago, I was at a conference in Vegas, and we were sitting around talking. I was having a conversation with one of the attendees, and she gave me two pieces of advice. One: "You're constantly overestimating what my team can do." And two: "Just talk to me like I'm your friend."

We're data wonks, man. We love data. We wake up every morning thinking about data. Our CEO, Mike Capone, often talks about how we believe data can fundamentally change not just how businesses and organizations operate, but really make the world a better place. And that's what we're all about. We’re obsessed with data and finding ways to leverage it for the better.

Chris Powell: I fully agree. We cannot achieve the full promise of AI without having the right data foundation.

Michael Krigsman: Sarvant, you have a vision for the transformative power of AI and data. That's very clear. Where is this all going in the future?

Sarvant Singh: Vehicle complexity will continue to increase. It may happen due to electrification or autonomy. There'll be more computing on the edge, particularly for a business with physically distributed operations like ours. 

So, I visualize all of this becoming more and more complex. And our goal is to continue to utilize disruptive technologies like AI and generative AI to deliver a great customer and associate experience.

Trucks are data centers on wheels: that is what we have been saying lately. There's just so much compute and processing going on there; it's unbelievable. We are sitting on more than two petabytes of data, and the pace at which we are accumulating this data is accelerating. We collect billions of data points on a monthly basis. On any day, we are processing more than 300 million IoT messages.

Michael Krigsman: That's incredible.

Chris Powell: Yeah.

Michael Krigsman: Chris, we hear how mature Penske is with their data and AI. What are other companies that you speak with doing to transform their businesses with data and AI?

Chris Powell: There are so many organizations, as you can imagine. Everybody's trying to figure out how to transform with it. Whether it's healthcare, banking, or manufacturing, we're also really proud to be working with a number of not-for-profits, charities, and non-governmental organizations around the world. Everyone is looking to see how they can better accomplish their mission with data, for sure.

On the larger side, take a company like Honda, which has been able to reduce their overall manufacturing planning process from 30 days down to one day. That change has had huge benefits in terms of their supply chain management and understanding what they need to be doing on a regular basis.

And then we work with much smaller organizations. Appalachian Health Care is one that I love talking about. It's a small organization. I think they've got about 15 different healthcare centers in the eastern part of the US, where they took a common problem (patients not showing up for their visits, which we all experience day to day) and used artificial intelligence and predictive analytics to understand who's most likely not to show up and try to drive a different outcome. 

They were able to raise their top-line revenue by six million dollars over two years. That's a huge difference for an organization of that size.

And then from an NGO standpoint, at Qlik Connect, we had someone from MedAir (Heidi from MedAir) talk about what they're doing. They're a charitable organization that works in some of the most difficult environments around the world to bring medicine, food, and supplies to people in need. If you bring data into a life-and-death situation, we often talk about things that seem like exaggerations, but they're really moving food supplies into areas that truly are life and death, and medical supplies that truly are life and death. And they're now able to use speed to collect data for things that used to be a report every two weeks; now it happens every two hours. 

So they're literally saving lives by leveraging data and really thinking differently about how data can change outcomes.

Key Strategies for AI and Data Adoption

Michael Krigsman: Sarvant, what can companies do to encourage the adoption of data-driven tools and the adoption of AI, much as your organization has done?

Sarvant Singh: First, start with a robust data foundation, because there is no good AI without good data. Second, make the tools easily accessible. Self-service is really the key, and we have to press the easy button on those things so our business community, our analysts and our citizen developers, can access data and tools easily. That's number two.

Third, build an ecosystem of partners. I may have mentioned this earlier, but I cannot overemphasize its importance. And last but not least, talent is the key. At the end of the day, AI, like everything else, is a people business.

Michael Krigsman: Chris, as we finish up, what advice would you offer to other companies looking to emulate the kind of transformation undertaken by Penske?

Chris Powell: First, recognize that you have many different needs. Understanding how all of these things can complement each other and working with a set of partners in that ecosystem is exactly right, as Sarvant said.

Second, we make sure to provide the flexibility that organizations need. As different pieces of technology change, organizations need to maintain that flexibility, whether due to a merger or acquisition or a change in strategic direction. They need flexibility within their environment to build these tools and capabilities.

Third is guidance. You need tools that help your organization move forward by providing guidance, and you need to work with people who can provide the support structure that you need.

Michael Krigsman: People and technology coming together.

Chris Powell: Yeah.

Sarvant Singh: Absolutely.

Michael Krigsman: Sarvant Singh, VP of Data and Emerging Digital Solutions at Penske, and Chris Powell, Chief Marketing Officer of Qlik. Thank you so much for taking time to speak with us; I really appreciate it.

Sarvant Singh: Thank you. Pleasure being here.

Chris Powell: Thanks.