From Pilots to Paradigm Shift: AI Transformation at Jabil

Speaker A: Welcome to the Executive Insights Podcast brought to you by aws, where we address vital questions and share unique perspectives from leaders at the intersection of business and, uh, technology.

Speaker B: Welcome to the Executive Insights Podcast brought to you by aws. My name is Taimur Rashid. I'm the Managing Director of the Generative AI Innovation Center. I am thrilled to be here with Vivian sun, who is the Senior Director of Data and AI at Jabil. Welcome to the show, Vivian.

Speaker C: Thank you, Tema.

Speaker B: So, Vivian, to get us started, you are leading AI transformation for Jabil, which is a global manufacturing company with over 140,000 employees across 25 different countries. What does AI transformation look like and how did you all get started with that?

Speaker C: It's definitely a journey. I think I probably should write a book about it. But before we dive into that, I'd like to give a little bit context of Djabo chaebol operate in 25 countries, as you were saying, with 100 factories. So we have a very large footprint and we also serve over 400 customers in a very diverse set of markets, ranging from consumer devices all the way to healthcare, from server and management all the way to AI infrastructure. So, uh, for chaebol innovation, entrepreneurial energy is always being there, and it's very important to navigate that to the right direction to create benefits and value. Uh, now going back to AI dribble, uh, actually started to utilizing AI about seven to eight years ago, utilizing AI computer vision, because that is a, uh, technology that's very benefiting us and we learn from that. And fast forward into recent years where generative AI is now bringing values. We're applying that model and expanding it because generative AI is not only giving us impact for operation, but also for finance, for business units and so on. So we are expanding that governance party to help the whole company. Now.

Speaker B: That's fascinating. And can you, can you talk a little bit about the point where you went from realizing that, hey, AI is not just a technology for these different use cases that you're doing across the company, but now it's actually a transformative catalyst for how things can be done m differently, how the organization needs to think about new application development differently. Can you talk a little bit about the impact of that? And then how did you all sort of translate that into action?

Speaker C: Yes. So there is a lot talk about whether AI is going to be a transformational effort. Is it going to be a paradigm shift? I actually believe it's going to become a paradigm shift because it's going to change everything in the enterprise and it's going to change how we work, how we live. So there are, uh, several things I think our AI brings very differently comparing to the innovations and transformations we had before. For example, Internet technology we had, we have transformed in the past. It took us one to two decades to come to when Internet becomes a utility. But AI changes every day. Today there could be an innovation, tomorrow it will probably become a table stake. So it requires us, everybody as a leader, as, uh, people practicing AI to learn about it every day. So that's a very big difference. Another difference is the demand for value generation because there is a lot investment being put fueled into AI and there is a mass belief that AI is going to generate tons of values. But people are expecting that. So when we look at a recent publication by MIT which says that they have used iceberg, uh, to explain the AI value. So what we are seeing now today is only the tip of the iceberg, which is the 10%, and there's still the rest of the 90% under the water, which that means we have to be chasing for that value very quickly and digging that out. Right. The third, I think is the most important is how we are educating people about using AI. Unlike, uh, the previous transformation, for example, when Ford introduced the car model T, it was a big transformation for everybody. But you learn how to drive, right? And then you can drive and Internet, uh, you take lessons until you know how to browse the Internet, find the information you want. But AI is much more than that. It's about learning how to live again, how to live with something never happened again. It's like learning a language, not a skill set. And so you can practice that language and continuously learning about it. So it's a progress. So I think these are very big differences that AI are bringing to us that we need to continue learn and how to incorporate that into the enterprises.

Speaker B: Yeah. Now, you said three things that resonate very well with me too. Which is, number one, the space is evolving very fast, which in and itself is very overwhelming because you have to keep up with all the underlying changes. And then secondly, we are truly seeing the tip of the iceberg, which is in one way very, very fascinating, but also a little bit unpredictable too. It's like you don't know what's below that. Right. And I think the third point that you mentioned around people, it's about people. And your CIO may yap, uh, talked about taking an employee m first approach to this transformation. How does that actually manifest in practice, in your experience as you drive that transformation at Jabil Yeah, I think a

Speaker C: big theme into what May has said is around empowering people because we cannot innovate for people. We have to enable them to innovate because AI is going to probably eliminate some of the job roles, but it will also create many more so leaders. Our responsibility is to empower them to transform themselves using the technology. So we want them to innovate and put them in the center of creating values so they can feel the benefit of doing that and are willing to continuously uh, learn and going forward on the path.

Speaker B: Yeah, no, and I think you're bringing up a very good point that it's very much something that we as an organization have to practice. And I think one of the important things is, and as you said with that employee first approach, how do you empower people? The great thing about AI is that it's not only a great equalizer in that many people can now participate in it, but at the same time it's a great multiplier. And so the only way you can get the multiplication effect is by making it part of the daily habit. And I think one of the big things there is around the change management part of it. Are there specific use cases in your experience at Jabil where you've actually seen the impact in the value it's driven or in the change of culture that's resulted from using those AI tools?

Speaker C: Many, many. Because we started uh, implementing AI quite a long time ago, I mentioned about AI computer vision. So we in the process of giving that tool to people, we concentrated on how to support them, developing their own solutions. Of course, when we started to test it out, building our prototypes of our solutions, we had ideas of uh, from the leadership from it, from uh, other functional teams of what are the, some of the benefit we can capture. But it's very, very limited. But after we designed the first phase of our prototype, we handed over to our employees and they have gone so creative in looking for the value using the tools. We started out from using probably 20 prototypes. We send them to our people and today has been expanded to over a thousand just in year of time. Right. And there are many, many other examples too. Right. AI computer vision is relatively more mature and we are also utilizing generative AI and agentic AI today. And when we look at on the shop floor, it is a nightmare when we have line down and when that happens the pressure is very intense. And we have operators, we have engineers, sometimes operation managers surrounding the machine, trying to figure out what to do. And instead of uh, flipping through the menus and trying to find the root cause and trying to remember what happened last time to today. We use AI to generate that automatically and instantly sending the information to them. So instead of uh, chasing for data information, they are now spending more time trying to find the root causes and take actions with much faster speed. And people love it. Not only we expedite the problem solving, we remove the anxiety from that such high pressure moment.

Speaker B: Yeah, well, and then the impact is not only in the efficiency gains, but then also to some extent in the employee satisfaction with their jobs. Right. Because you're eliminating what normally was a very cumbersome and probably hard thing to do, which is looking for information.

Speaker C: Right.

Speaker B: You know, with AI being so new, this culture of experimentation is what's needed to really see what's the art of the possible. But that obviously comes with that fear of failure. And oftentimes people have that issue with okay, I'm going to do something, I'm going to fail. Right. But culturally, how do you create that safe space within Jabil so that you not only encourage experimentation, but you also look at it in a way where like failure is not wrong.

Speaker C: Exactly.

Speaker B: It's just part of the journey.

Speaker C: Exactly. And it is a mindset shift. And we uh, are very conscious on developing that. Developing AI is a little bit different than traditional IT applications. Since I come from the IT world, it is important to emphasize or develop that culture of win fast, fail fast. Because AI values are accumulated with a steady, uh, iterations of development. So we emphasize on um, people watching the value at every iteration instead of trying to build a very monolithic or perfect solution, watching and tracking the value at each iteration and concentrate on continuous improvement. And from that organizational perspective, because chaebol again is very decentralized, we're very diversified and it is very important to balance the governance versus innovation. And so we actually built a, uh, federated operating model. And referring to the paper you wrote, uh, octopus organization, that's exactly what we're trying to practice. And we want to, the central body in the middle will be responsible of building the strategies, giving the directions, building the guide rails. Right. But we want to give enough freedom and autonomy to the little heads on the tentacles like the octopus. Right. So tentacles will be able to drive in the directions that best for them. Right. With that combination of uh, octopus organization in a federated model. So we can find a very good balance between innovation and aligning to overall strategy and guide rails. And I think the octopus organization is enable a good balance of innovation autonomy versus policy in the central head. Uh, if I Can say because the central head is controlling the strategy, controlling vital systems in an octopus and strategy directions in an organization. If we were to translate that. But I talked about two way communication between the big head to the tentacle heads and tentacle hats, feeding information back to the big head. Right. So of course that's going to enable the agility and align that with the strategy, uh, and policies. But I think it's going to, if we practice that well and in an organization, it's going to replicate the intelligence of an octopus. We know octopus is a very intelligent animal. If we translate that into an organization, then we will probably truly translate, transform our organization into a very intelligent organization. So I think we have witnessed some of that in practice and that was a, ah, nice surprise to me.

Speaker B: That's fascinating.

Speaker A: We hope you're enjoying this discussion. To join the conversation and engage with other business leaders on these topics, follow AWS Executive Insights on LinkedIn.

Speaker B: And as we think about that overall transformation that's happening within the organization to make it more intelligent, one of the things that we often think about is how do roles transform? What are those new skills of the future that not only the existing workforce has to develop and build, but at the same time as you think about bringing new talent into the organization, how has Jabil thought about what those skills are and what's required for the future workforce?

Speaker C: I think it's important to be uh, learning, individual learning organization because today we have AI. AI by itself is iterating and we will be facing other new technologies. Right. When talking about AI, it has transformed from AI computer vision all the way to now we're talking about physical AI. And so being able to learn is very important from the workforces. And it also gives a lot of challenges and opportunities for our leadership as well. Because today our leaders are managing people, tomorrow our leaders will be managing AI and the people. So it is very important to understand how to perceive AI. Just like how we are perceiving people today. How do we give AI the work that it is best at and how to integrate that AI workforce with human workforce. So there is a lot of additional considerations that our leaders will have to do and how do we train AI as well. So those are very interesting topics and we're also exploring and learning and hopefully uh, we can get mature step by step.

Speaker B: And you can't start off with the right solution from the beginning, but the most important thing is to start somewhere right?

Speaker C: And we have to realize that the governance itself is transforming because nobody has experienced the governance in the AI. World as leaders, we also, not only that, we give the room for our employees to make smaller mistakes, taking smaller risks. We should also be doing that for ourselves. We start, let's start from somewhere. But keep in mind very consciously it is a continuous improvement as we go forward. We can perfect that governance model and improve that along the way.

Speaker B: Yeah. It truly puts meaning to the mantra that you're building the plane as you fly it.

Speaker C: Yes.

Speaker B: You know, one thing I wanted to talk about, you mentioned the shop floor and how you enabled AI for the shop floor workers, uh, obviously giving them much easier access to information where now they can spend that time gained now towards higher level sets of activities. Right. Can you talk a little bit about that? And how should leaders think about using efficiency gains in one area to either doing more of the same so they get better throughput, or maybe looking at new ways of doing things?

Speaker C: Right. Definitely it's going to be a combination of those. We have uh, many use cases that we have seen, um, people successfully utilizing AI to remove repetitive non value add work. But we should not stop there. So using AI is not only to do same thing faster, it is also to leave the room for more valuable things. Some examples I can give. We often have to hold those KPI reviews. We want to make sure every day when we deliver the product out of the door, it is following meeting the KPI. There are a lot of meetings on our factory to review the KPI regularly and oftentimes managers get into one room and they're trying to understand from the dashboard where the information is when they see something underperforming and they want to find the data behind that number before they had to either ask it to extract the data from the servers and building additional dashboards, which is very time consuming. Today we're utilizing AI to do that and the AI can even prepare for information before they go into the room. So that expert is not repeating what the managers were doing, but it's giving them more time to analyze the problem and making decisions instead of hunting for data.

Speaker B: How do you think about foundational things that need to be established so that any department that wanted to activate AI, uh, for their line of business, they can do that in a scalable, consistent, secure way.

Speaker C: I want to explain that into again from different aspects. First aspect is probably from technology. There are so many kinds of technologies out there and we constantly spend time to think about buy versus build, what should we buy, what should we build, if we build what kind of platforms we should be building. So the JBO strategy on technology Stack is to go with big partners like AWS who surely will be very successful in the next 10 years. And we try to make that standard across the board so we can start from standard technology to help us proliferate more. Because Djibo is not a technical company, we don't want to build AI, uh, models, we want to utilize it. And uh, that's one aspect. And at the same time we want to utilize what everybody else is building. That comes to the buy strategy where for example, we want to understand what SAP is doing in the AI space, Workday, ServiceNow, Snowflake, so on, so we can tap into those development as well. So that's from the technology perspective and from human perspective. We want to make sure we can find tools and different kinds of tools to fit different business needs and to fit different business knowledge on technology. So for example, we utilize Amazon Quick Suite. There is, um, Amazon quicksuite Flow and there is Automate. Right. So Flow probably will become a tool that our business people use. Use because it's easy to use, it's very quick to build, uh, small use cases. But Automate probably is more of it or more experienced user to use. Right. And organizationally we talked about octopus organization and we need to also expand that to different level. We need to get attention from our C levels and from the middle management as well as um, everybody else in the organization and making sure we have very good communication channels and continuously educating people.

Speaker B: Yeah. And that's fascinating because as we think about transformation, it's not a beginning and an end state. It's the journey to get the transformation right. Where on one hand you have culture of experimentation, acceptability, that, hey, failure is part of the journey, but that's part of the learning. And then as you do that and you go from one use case to the next use case, you have best practices, reusable assets, reusable approaches. And then as you do that over a period of time and in a very consistent way, you are living the transformation and you're changing as an organization.

Speaker C: Yes.

Speaker B: If you could go back in time and for the leaders that are either starting their transformations or in the middle of their transformations, what are one or two pieces of advice that you would give now that you're kind of living through that transformation?

Speaker C: So if I'm looking at how we have come to today, I would have been a little bit more insisting on, um, solving business problems. Right. And put business value in front of technology and then combine those two, business value and technology feasibility together to at the very beginning of the use case, uh, of building the solutions. And that's very important. And the second aspect is not to underestimate integrating AI to the legacy IT systems, because generative AI is going to be integrated into our business processes, integrated into our applications, our infrastructure. And uh, these infrastructure applications have been living in the world for decades and plan for that upfront and understand how we can streamline the current process, the current infrastructure, so we can adapt AI better because it's going to cost us time for doing that. So it's better to think about, uh, that upfront. And like every transformation is organization, change that evolves process technology and people. It is continuous process. Don't give up. It's going to be, uh, difficult. But find that lower hanging fruit that's going to, uh, stimulate the further appetite for change. Right? That's super important. And from there build the momentum. When it comes to the certain point, it will start to replicate by itself. That's when it comes to the sweet spot of transformation.

Speaker B: That's very true. Yes, very true. Well, Vivian, thank you so much for the insights that you've shared and thank you for joining us with the Executive Insights podcast.

Speaker A: Thanks for listening to welcome to this episode of Executive Insights, brought to you by aws. If you enjoyed this episode, help us spread the word by rating and reviewing. And if you haven't already, be sure to subscribe so you don't miss an episode.