00:01:28
And yeah, I mean, where do I start? So I kind of view technologies and I've always kind of in the past few years, I've always viewed technologies from the kind of lens of operations and what I call operations is manufacturing, not what a lot of tech companies refer to as like AI ML operations, because that's just stealing the words from manufacturing, which so much of tech has done. And kind of what got me here was kind of a long path, I would say, and this is where the long part comes is, you know, coming from an engineering background, studying mechanical engineering in Scotland, I've kind of took an interest into, well, how do you bring sports engineering and mechanical engineering into the same realm?
00:02:15
And that's when I kind of set about avoiding the oil industry where I was destined to go, because that's kind of where I'm from, originally in Aberdeen, we call it, which is oil and gas, and kind of went on a journey and trying to find how do I get into companies like Adidas? And that's when I joined Adidas after university into a graduate program and was kind of working in their product development area, figuring out, well, how do you, you know, how can we transform a football boot to make players jump higher, faster, kick a ball harder, et cetera, and quickly realized that that's kind of a lot of novelty and a lot of marketing kind of into it.
00:02:55
So after about a year and a half in Germany, I was offered the opportunity to go out to China to work in the manufacturing there. So I went, yeah, sure, why not? Probably a bad decision at the time. So I moved out to Guangzhou in China for four years, and there I was heading up with Adidas, what they called operational excellence teams for the specifically focused on automation and technology. And we were exploring how we could kind of drag the footwear manufacturing industries by, you know, by the heels into the 21st century, into a more advanced manufacturing state, which was a very kind of, it was kind of a new concept for that industry, because they're very used to moving countries to take advantage of lower labor costs.
00:03:45
And what we were trying to do is say, how do we use technologies to avoid the move and actually enable a higher levels and more efficient manufacturing whilst maintaining the same kind of level of quality? And that was a really interesting time. We were introducing a lot of new technologies, changing the mindsets, the culture in these kind of quite large industry, but quite like a rudimentary industry, let's say. And during that time, we were going through, how do we introduce machinery? How do we change processes? How do we, you know, look at chemicals to do, to make things more efficient, more effective? And that was when I started to really push the narrative of, well, where does digitization come into this picture?
00:04:24
And obviously, footwear being a little bit further behind electronics and automotive manufacturing, we had a lot of industries to look up to. And that's where then I decided to extend my time in Asia and move to Vietnam for another four years. And that's where I was leading for, you know, all of the factories across the address in Asia, 65 factories, I was leading their digital transformation journey. And that was, again, really driving forward, like, how do we get into the mindset of adopting these technologies? How do we build our digital footprint? How do we go paperless in the production lines? How do we get data from critical equipment? How do we actually then start to leverage it in an intelligent way? And again, we made a lot of progress in that over the four years that we were doing that.
00:05:12
And it was really fun running hackathons, which was the first thing, you know, first time they had done anything like that and really driving that transformation from the factory level up, which is always the best way to go. So when I talk about operational AI, that's for me is the definition is really how do you go from operations up to AI and work up to that? After spending eight years out in Asia, I decided it's probably time to come back to the Europe. So I went and did an MBA in Barcelona, a university called the ESA, which is well respected for kind of ethical business. And that was a great time.
00:05:50
I call that my mid-career life crisis where I went out, I spent a couple of years in Barcelona, did an internship there and got introduced into the world of technology even further. I really wanted to explore then how large companies who are more advanced in digital manufacturing were adopting more advanced technologies like AI. And that's kind of where then I was led into my last company, which is called C3 AI, which is where Cedric and I met. And that was really interesting because they are really doing large scale deployments of AI, you know, in large companies like Shell or even, you know, Baker Hughes or some other like chemicals and food manufacturing. And that was really the other spectrum side of the spectrum as to what we were doing at that.
00:06:32
So that was a great insight. And I've continued on this work now into my most recent company. It's called SoftServe. I could probably talk about SoftServe a lot more in another day, but I'll stop this long introduction for now. I mean, it was an expansive introduction, but I already learned so much more. I can't believe like you spent eight years in Asia. Before we go into the digital transformation, because recently I went to the Middle Americas or like South America, depending on who you might ask, like Ecuador, Colombia, where they were also, let's say, adopting our digital tools. But what I noticed there, especially in comparison with, let's say, in European countries from time to time, like the discipline to execute on day-to-day tasks and like how they were really looking towards technology was a lot different there than it is in Europe.
00:07:29
So I'm really curious, like what was for you the biggest difference going from Europe to Asia and China? And maybe there is even a difference between China and Vietnam. Yeah, certainly. And I think that varies from culture to culture, industry to industry, even company to company, you know, how you want to go about introducing technologies, I think is really interesting within any sort of organisation, culture, whatever. I think it also is down to you how you get people to engage, looking for those quick wins, you know, at Adidas, we were still in kind of the infancies of what we were doing. So we were kind of testing and playing with a lot of things. What we were doing was introducing, you know, Raspberry Pi into the conversation.
00:08:15
How can we throw sensors here? My favourite part of my job was hiring interns from the local universities and just saying like, hey, guys, here's a Raspberry Pi, go like attach some sensors to this machine and see what we can see, what we can do. And that was the funniest part of the job, to be honest. I talk about the hackathon, that was also super fun. And it's like and within those cultures, if you talk about China, it's like one of the fastest moving places you can you can work. I mean, I saw buildings going up in three months, you know, 50 storey high, you know, storey buildings. And they have a culture of just doing everything now, which is like super fun. And they work six days a week and they can execute stuff super quick.
00:08:59
They might sometimes do it wrong the first time, but, you know, because they can do it so fast, they'll just get it right the next time. And I really miss that culture since I've arrived back in Europe, where it's a much slower, kind of like methodical, like, you know, we have to think carefully and to set up some projects and meet several times that you get something going. Whereas and I miss also being on production floors because production floors are kind of messy, chaotic. And I think that approach to how you also can innovate within the digital space is also super. It's also super important is test, learn and retry. And you can do that sometimes with low cost technologies like Raspberry Pi and then evolve it into more, more mature things later.
00:09:41
And I think that's what I enjoyed about some of the stuff out there. And again, in Vietnam, they have such a young age, average age of the country. And they have a really strong digital, like they're really strong in the kind of programming and they have Internet everywhere across the country. So they have a really strong base of like young engineers who can build and test stuff. And we would partner with the local universities to do that. And that was always the funnest part of the job. I thought interesting. I never heard about it or I didn't know that Vietnam is really this like a way forward with the digital strategy and the way forward. I have also a question about your career, because it's definitely comprehensive from your introduction.
00:10:25
But what is your what do you consider as the biggest achievement in digital transformation so far until today's day? Well, I think the work we were doing at Adidas was significant. I mean, really going from nothing as a kind of strategy to really developing that into a full fledged like roadmap that we had built out for. And it's still running today, which I'm quite proud to say I caught up with some of the colleagues not so long ago. And they're still working on some of the stuff that we had prepared and had said, OK, this is what we should do. Even after I was leaving to do the MBA, you know, this is the this is the roadmap that we have to build out.
00:11:04
And I think the great thing about that was more the kind of ecosystem that we had to build. And as I mentioned before, we were looking at other industries and saying, well, you know, how does this work in our industry? The problem about footwear or, you know, making footwear, as I say, is the hardest thing to automate in the world, because I say automating an iPhone or a car is easy. But with a shoe, you have a left or right 12 different sizes and the colours and materials are changing all the time and you can't get it into the right position. And even for one shoe, it can vary in like one to two, three millimetres. So always try the shoes on in the shop and try another one if it's not fitting, I say.
00:11:43
But it's a hard thing to automate. And that also then means that when you're looking at a digital transformation strategy and you're looking at it for your company or even your industry or even the location you are, you have to customize it. Like you have to really think, OK, well, that's great that those guys are doing that, but that doesn't really work for us. And we had to start with that kind of initial like initial kind of research and understanding what is available to us. And we went about testing and using technologies and using the ecosystem of suppliers that we had in Asia to figure out what worked for us. And we did a lot of really cool stuff. We were looking at, you know, AI vision systems. We were looking at AGVs.
00:12:21
We're looking at robotic systems. We were doing quite a lot that was not just for the sake of automation, but also for the sake of sustainability, saving a lot of materials, saving a lot of glue, avoiding help, helping keep people away from dangerous processes. And that for me was like the scale of what we were doing across the 60 odd factories that we were working with was significant. And I was really proud of that work that we did. And to hear that it's still going today, it's just great to have a legacy, I'd say. Yeah, I think it's always impressive when you see that your legacy still lives on after you, right? Like those are also usually the things that I look after, like that even when you leave somewhere in a role that there's the legacy that you leave behind and that it just continues.
00:13:11
Some things might change, but at least the North Star to what they go to is still the same. And I think what triggered me as well was like in Asia and China, like the mindset is more of action. And I was just reflecting on that because I think Marie would really fit well in that kind of culture where like you just start and like when something is asked, like you just start and start taking action. So Marie, maybe Asia for us might be a good destination because you like action a lot as well. What I always liked when I was in China was, you know, people always say like in China, they're copying other people's stuff or faking other people's stuff.
00:13:52
What they sometimes refer to as a re-innovation where they just re-innovate on top of what other people have done. And by the time someone tries to come and contest what they've done, they're already five iterations ahead. Which I always think is a very novel, I mean, interesting approach. You could question the ethics of it, but it moves faster, right? Definitely. And I recognize that as well. I was actually curious about, let's say, that journey that you've been on, especially with the notion on AI and artificial intelligence in digital transformation in operations, in manufacturing areas. Like, how do you see that AI is effectively having an impact there? Like, what are the tangible examples where artificial intelligence is really adding onto the manufacturing lifecycle?
00:14:53
Yeah, well, again, looking back to the title AI operations, operational AI, what I like about AI is it has the potential to enable operations teams to behave fundamentally different. So we know that AI enables us to do essentially three things. Predict, optimize, and prescribe. Which I have coined the phrase in the past two weeks, POP. Whether this catches on or not, we'll see. I hope it will. Yeah, exactly. But for me, the first part, which is predict, if you look at operations, operations teams are always firefighting. You're always catching up with what happened yesterday or two days or a week ago. What we want to do in operations is, you know, it may sound funny, but they probably, as much as they love being in the firefighting phase, they desire a very relaxed and chilled lifestyle.
00:15:53
You know, they'd love to set up operations and just let it run. That's the dream. So what predicting enables you to do and what AI enables you to do is look through all of your historical data and be able to recognize patterns and be able to then predict what is coming your way. So you're able to look down the line into the future and say, in two to three days time, this machine might break. In two to three times, I need to manufacture this much product in a week's time or a month's time. So that is where AI comes into operations to transition you from a reactive state to a proactive state. And once you've made that prediction, then AI can then step in again and enable you to optimize the outcome.
00:16:34
So you can say, well, you know, how much do I need to do? What's the optimal settings that I need to apply to this machine to achieve the outcome I need? What is the amount of inventory I need to hold to achieve the desired outcome? And that's then the optimization phase. The final one is even better because you're starting to not rely on the knowledge experts within the organization. You can capture the people who are experts within industries, manufacturing, and you can capture that knowledge and then use the AI to prescribe the knowledge. So, OK, this machine is going to break down in five days time. This is what you need to do to optimize the outcome. And this is what the actions are that you need to take and when you need to take them.
00:17:19
So that's that prescription phase. So that prediction, the optimizing prescription is super critical. And you can apply that across operations, whether it be supply chamber, you are Cedric. It could be within the manufacturing. It could be for working with machines. It could also be, you know, how you do production scheduling, for example. So there's a large variety of places that you can actually apply that methodology. And that's where I think it becomes quite interesting because you're changing the way that operations then starts to behave. I have a question regarding your current position with SoftServe. Is it that you are doing consulting to the different companies that it's more like variety of the companies that actually are interested in the industry solutions? Or how does this work? So technically, SoftServe are an IT consultancy.
00:18:10
I didn't really know that before I joined because I've always been kind of against consultancy. I always say it's con and insulting put together, consultancy. But yes, technically, as a consultancy, the way that I see myself is in my role is I'm able to speak the language of the operational teams and understand the problem statements and the value that can be extracted from what they need to achieve. And then also be able to help translate that into what an IT strategy or IT solutions could bring. For me, the process is quite clear, though, that we do not start with, you know, a hammer trying to find a nail process and that we have a technology and we're trying to find a problem. We really start with the problem first.
00:18:57
We define the problem. We understand if it's actually going to add value to the operations and we then work our way backwards and see what is the technologies that are relevant or right to actually solve this problem. And then we can actually start to then build those and do it in a, you know, we talk about agile fashion like we did in China. How do we get a POC up and going or an MVP to then to get that resolved? And that's where I kind of see myself as not consultancy, not IT, OT, OT, OT, IT conversions, I think is often what it's referred to. The main reason I ask is like, what do you usually apply on your clients or the companies when they come to you that they want to do the digital transformation of a specific issue?
00:19:45
Is there some method or something that you do as the first, you know, like they come a bit to you that they want to do the digital transformation? What is the first thing that you usually do when they come to something practical? Well, probably first of all, understand what the strategical goal is from the top of the organization is that, you know, are we trying to improve our delivery and set customer satisfaction? Okay, well, then we work backwards from that or work down from that. Is it that we want to increase the capacities of our facilities? Is it we want to reduce our, you know, our cash flow or like reduce our free up free up cash flow? There can be a multiple, multiple statements that we're trying to look at or problem statements we're trying to address.
00:20:34
Part of that process is obviously going through that conversation. Some people might have some idea as to what they're trying to do, but they might not fully understand why they're doing it or what they're trying to achieve. And it's being and kind of this is, again, part of my role is being within the operational mindset. We can really be that sparring partner to say, is this actually what you're trying to do? Maybe you'd be better off doing this. And we can also bring in examples of what other people are doing or what we have done with other people and say, look, well, this is a kind of a low hanging fruit and this will likely return this sort of result. And this can be scaled up across these number of facilities or areas, for example.
00:21:16
So there's different approaches that you probably have to take with different people. Some people just might just want to do what they want to do. And that's probably not the best people to work with because they're not going to be very collaborative. Cedric, I mean, you're probably in this realm as well where you're working with partners and consultancies in your role. Do you think you're always right? I mean, there's experts in the field that we can rely on. I think what I always look for in people that I work with is more like, are they able to form the bridge? You know, I don't look necessarily at expertise in one field, but I'm looking at people that are able to make the connection of like, OK, the operational reality to the, let's say, then the technical requirements or the opposite way, right?
00:22:15
Like being an expert in the technical field, but still being able to connect that to, OK, it wasn't the business implication of that. And that's where I liked your approach where, and I think your operational background helps you there, right? That you have the operational awareness and the things that actually matter to someone who's on the shop floor, someone who's managing the shop floor, and then help translate that kind of business statement to something that an IT developer or OT developer can actually work with. And that's what I also look at when either working with consultants or internally with people, like, are they able to make that bridge? For me, that's going to be like an increasingly important skill to have in the age of digital and in the age of AI, because we will get to know less and less of sometimes the details of the technology, but we still need to make that bridge.
00:23:15
100%, 100% correct. And really, I'm very reluctant to get started on a project unless I've actually spoken to the people who are at the goal phase, who are actually having to, will eventually use that technology, and are the people with the real problem statement. So, you know, the reality of when you start talking to someone who is performing a maintenance task or actually doing the scheduling for a production or, you know, running a machine, when you actually go talk to them directly, it's a very different picture that gets painted to what you might have been communicating from like a middle level or a cross-functional team. And so, for me, it's always critical to go talk to those people.
00:23:55
And, you know, when I was living and working in China, like, I made my pretty good effort to learn Mandarin so that we could go talk to, because the best part was doing gimbal walks and walking through the factories and being able to question and talk to the people on the production line. My Chinese got to a good enough level to do that. That's not like, not into some depth. The highlight of my Chinese Mandarin was talking about Scottish independence with a taxi driver, I will say. I don't know how I did it, but I think I'd had a few beers. Nice, very cool. But I think grounding any assumptions to the shop floor is always critical before you get started with the project, because then you're not really taking a true agile approach.
00:24:42
You're not delivering, you're not going to end up delivering the results that are necessary. No, definitely. And I also really liked the approach of, let's say, or like the definition of AI is like helping us to go from, indeed, the reactive mode. Because, I mean, I see it here every day as well. Like, there's fires to fight almost every day, almost every minute. But to go to leverage AI to make us go from that reactive state to a more proactive state. And I'm wondering, like, because I see it here, we still have a lot of leaps to make. But where would you see that? What's the evolution that you've seen in that sense with the companies that you have worked with?
00:25:24
Have we really made a step up in industries or are we still where we were in that sense like five years ago? I'd say a lot of companies that are even more advanced are still in their infancy. And I mean, I think a lot of people, and this gets thrown around a lot, technology, right? I always say technology probably is 30% of the equation. Even if you predict something and tell someone or a team of people that something has to be done, whether there's the systems in place and the culture in place to actually then go action, that is a very different thing. Because that is what is also important. Are they incentivized properly to go take action upon what is being prescribed and done?
00:26:14
And is there a culture to believe in the systems that they're working with? And then is there actually a process flow and etc? There's a lot that goes into actioning what an AI system might provide. I mean, there are areas of AI where you can apply it directly. It's like AI vision system cameras and it's fully linked up in a closed loop system to a robotic arm or something else. Then there's no human in the loop. But these systems will have to operate with humans and there will have to be kind of a back and forth. We don't necessarily want to rely fully on AI to do everything for us. We want it to be kind of like a sparring partner again into how we work. And there will be an open loop.
00:27:04
Most of it will be open loop advising and working with someone to get the results they need. I would like to speak about the risks of AI as well in the operations. In operation, which is also maybe goes hand in hand with reactive versus proactive. Are there any risks when we look at the operational AI where the people are, for example, really like leaving AI to automate things without double checking of a human? Or are there any risks for the current operations that we see that can be in the future really problematic? Yes. I mean, it's a very good question, especially in operations where safety is super critical. The outcome of what's being done is super critical. And I think, yeah, there is definitely a risk which has to be addressed, has to be managed.
00:28:00
With any risk is also usually an upside as well. You know, two sides of the coin. If you take a risk, you're expecting also some level of return. You can look at that probably. I mean, we're talking about AI now in two categories almost. You have your boring traditional AI, which is your regression models. And how do you predict? And now we have our kind of shiny new toy, which is generative AI. And I think both have kind of different risks associated with them. But I think that open loop factor that I talked about is definitely always a safety blanket to enable someone to apply like expert logic to what is going on. So if a system says like, you know, you should change the settings to 12 when really the max is 10.
00:28:54
There should be someone obviously to make logic of that. But that should be a very rare example. And that should also be then someone feeding back into the system to stop that from happening into the future. So there is definitely training. There's adoption. I mean, and this is where over the next few years we'll probably see quite a lot of people slating or trying to bring down AI. Because AI is, in my opinion, is still in its fairly early days. And I think laggards and other people are often very quick to try to kill something before it's really flourished into what it can be. I always say technology, bringing a technology or anything into any environment, manufacturing, etc. It's like raising a child. You know, you have to really foster it.
00:29:40
You have to let it mature. You have to let it make mistakes until it really develops into a very mature system. And you have to protect that child in that period of learning so that you let the right people test it who are more optimistic or in positions where they can have some risk without really facing then a more critical situation where that's going to kill the complete project. Because that is a digital transformation process where you're trying to make sure that there is success from introducing new technologies. And that is tricky. It is tricky, especially in large organizations where people are quick to say, no, that doesn't work. And then that quickly, when someone says that doesn't work, everyone else starts to reciprocate the same thing. And that's super dangerous.
00:30:25
Yeah. Failures reciprocate a lot faster than successes do. Exactly. Exactly. And it's a shame, really. And also how you are speaking about AI, it seems to me that the engineers are currently at guts because they are actually creating the AI and they're overseeing what the AI is capable of. How you were also explaining, it's like a child, you need to foster it, you need to take after it, what decision it does, how does it evolve, etc. That the engineers are becoming the new guts of the AI as such, because they are the parents of the AI, let's say. And the business just navigating the parent how they want to see it then in the long run. I mean, yeah. And I think what we feed the AI, I mean, we talk about AI models and we talk about their food.
00:31:22
Their food is data. And it's important that, same as a human. Human hears something, they repeat it. You teach a kid something, they repeat it. You're an AI and a human are very similar. And I think, and what I'm seeing generally across the data world at the moment is that, or across the digital transformation world at the moment, there's a lot of focus on AI models and LLMs and how do we bring these into production or into our operations and business. The conversation, I think, in a couple of years' time will be the adult in the room saying, well, what are we feeding these systems? And I think there will be then a refocus on, well, how do we get better data sets?
00:32:05
And how do we get the right data into these models that they're behaving better as well? And that goes back to your point about risk is that the better in, the better out. So we need to make sure that the models are being fed and trained properly as well. And are there any kind of known gaps in the data currently? Because, for example, what I know that AI, for example, in medicine, there is a gap that we don't have enough data on women. So how can we make AI to make decisions about the body of a human to make some process in the surgery? So when it comes to the operations, what are the gaps in data that might pose some questions or the risks currently?
00:32:54
What do you see in your daily workload, et cetera? So we're quite lucky in manufacturing because we get to talk to machines. Machines are fairly simple beings that we can add sensors to extract the data. The data is usually, well, behaves and the data is also customized to the operation or to that piece of equipment. So there's not as much kind of subjectivity or we can build the models fairly on based off those simple systems. I think where external data becomes more apparent is more in Cedric's world, which is in the supply chain, where you're trying to get better data or into the pharma world. We're trying to do the research is getting better data sets based on information that either is publicly available or that other corporations are producing. So you can buy data, for example. I have a friend at the moment who's actually working on a project to enable people to get better data. And I think it's going to be very interesting in a few years where that organization goes. And beyond that, then there is a lot of ways that you start to clean process the data.
00:34:14
And try to kind of expose data to different ways of behaving so that you can actually then get the right balance as you feed it into an AI model. I mean, what I see here or what I've seen quite often is machines, they're quite straightforward, except when you need to recalibrate them and they can have a little bit of a spike. But besides that, they should be running consistently. But it's usually at the data hubs, when it goes from the machine to, let's say, one system and then to another system and then to another. Where you add context to that data, but then somewhere along the roads, like I don't know how it sometimes happens, right? But like it goes into, let's say, your ERP system, but then it also goes to another system.
00:34:59
But because of different contextualization, all of a sudden, you start from the same data set, then you have two. Then you put those two data sets actually in your data lake. And then you're like, wait, it's the same data, but I'm actually getting different results. And then you need to have the data experts actually to really trim down, okay, where did we go wrong with that data set to then retrain the LLM. At least those are some things that I see in an operational environment where there really needs to be a lot of attention indeed now, but definitely in the future on what are we feeding these data sets to really understand the outcomes. Yeah, totally true.
00:35:39
I think this is then where we verged too far from operations into IT and I get totally lost and I have to rely on other people to make the responses. So I won't get into architectures and data lakes and all this and federated data and all this sort of stuff. I'll probably get lost as well. Yeah, exactly. But I mean, Cedric, does your organization, as much as you can state, do they rely quite a lot on external data as well as internal data? I think, I mean, majority I would say is internal, but we do rely for minor parts on external data like market trends and things like that, weather data. But I think majority of our data does come from internal sources.
00:36:32
But it's then like how internally we have manipulated the data in the past that I have seen that with, let's say, older systems that we have digressed from each other. And now when we're trying to jump on the AI train, we're trying to refit it into that it's one data set to then move forward and take the right conclusions out of it. I think the area I'm kind of most, been most excited about since I joined SoftServe as late has been really how we start to bring all of this stuff together. So with NVIDIA, like everyone obviously talks about the GPU capabilities, which are fantastic. We work a lot at SoftServe on the Omniverse capabilities. And the really cool stuff about that is that they're really creating an ecosystem there where you can bring all of these elements together.
00:37:24
So you can bring together AI models, large language models, CAD simulations. So you can take in, say, FlexSim, for example, and you can really start to then build and test out like physical simulations and virtual simulations all into one environment. And this is, I think, the next step is, you know, once we have good data, we have good models. How do we actually then kind of bring them together into one kind of, well, metaverse is one terminology for it. I'm not too convinced that that whole maybe is a good system. But what does that enable? And what some of the guys are working on in the robotics lab in SoftServe is really cool because they actually have some like physical robotics labs available in Ukraine and also in Poland.
00:38:12
They're actually working on how they can actually control, say, robotics with large language models. So they can say, OK, it's a large language model, you know, pick up this blue cup and build together with a robotic system and a vision system. The system can understand what is being requested to do and actually then go and execute the task. And this then, for me, really starts to open up like a world of opportunities within operations and manufacturing and how going from a large language model into like an AI model into then your actual physical operating systems becomes super interesting. And I think this is really the kind of evolution that's coming after where we are at the moment as well, is how did these kind of systems all synthesize together to create like a compounded result?
00:39:01
Yeah, and that's usually also the most interesting part because like everybody's always want, everybody wants to have like the end result when everything's already compounded. But right now we're trying to really fit in, like, how do we actually fit in the different parts? And we do it then by testing one part. But then, of course, it doesn't have the other part yet. It's not achieving the result that everybody's like super excited about. It's like trying to still use that momentum to keep them excited. And that's sometimes, I think, a challenge on its own, you know, like while you're building, keep people excited because like the end goal, we will get there, but we need time. Yeah, for sure. It's a long journey.
00:39:42
And as you said, like, I mean, especially in the supply chain where you work, those systems can compound. You know, you go from like demand forecasting is kind of, I always say demand forecasting is the start of the avalanche. If you get that wrong, the avalanche gets bigger, which means more poop, let's say, flows down the hill to the people. But if you get the demand forecast right, then the rest becomes kind of exponentially easier. And then once you get your inventories right, and once you get your production schedule right, like it is these technologies kind of impact each other. They're not running in silos. They're running in unison. And once those compound together on top of each other, you end up with what could be a significant result.
00:40:25
But I think you're right in saying that most of us are focused that, you know, how do we apply them within these silos at the moment? And prove the results and scale them out and see kind of how that all comes together. And once it's scaled, because scaling in itself is one of the hardest challenges. I have a question I would like to switch a bit from operational and to... No, let's stay in operational. And I would like to speak more about you, Scott. Tools that you are using on your daily basis, something that really facilitates your daily personal or professional life. Something, for example, in your laptop or something that recently really made your life more effective than it was before. Outside of the office suite, eh? Outside of PowerPoint and Excel.
00:41:19
So, yeah, obviously we use those tools quite significantly in our job, like Teams and Office and PowerPoint. Which, yes, I have to say I'm quite good on PowerPoint these days, which I'm ashamed to be. But we do eat our own dog food, I think. I think that's the right terminology. Is that we are using Gen AI internally to get hold of knowledge quicker. So we're using that and we're testing out ways to find information that's held within the organization. SoftServe is not a small company, it's 11,000 people. Yeah, it's bigger than you think. They've been around 31 years and based out of Ukraine originally. So there's a lot of interesting, especially at this moment, obviously it's an interesting company to be in. And it's a really fantastic place to be.
00:42:15
Reach out if you want to learn more about SoftServe. Hashtag, plug. Shameless self-promotion. And there's a lot of knowledge kind of divided around the company within some people's brains and sometimes in documents and sometimes here and there. And Gen AI really is a useful tool to help you kind of quickly understand where the information is. Be able to get it quickly and then be able to actually use it. And I think that's where Gen AI is doing a really good job and has the potential to do much more in tons of environments. We're seeing this in operations at the moment. I'm bringing you back to operations. We can use Gen AI to get closer to our data. Within manufacturing, you can use it to understand how to fix a machine faster, how to do a changeover, how to do a cleaning.
00:43:10
And I think that's really important because within manufacturing at the moment, we are seeing it's hard to get hold of resources. So we're seeing that the older, more experienced engineers that have been doing it for 30 years are leaving quite regularly and their knowledge is going with them. So what we can do is use Gen AI to capture that knowledge and be able to actually then help to bring the younger engineers faster up to the same operating capabilities as these guys that have been there doing that 30 years. And that for me is not like an ROI that you can put a number on necessarily. You could try, but it's really that self-benefit, which is producing a lot of results. But anyway, going back to other tools I use, what else am I using?
00:43:59
Well, I have another question, the last from my side, and it's a digital transformation in me triggers a lot about change. That you are actually changing the atmosphere in the company, that you do the digital transformation, you are enhancing the digital tools that the company is, for example, using. And I have a question about this change because change many times equals negative connotation in people because people just don't want to get used to new softwares, new tools, etc. So what is the most that people don't want to usually change into? Is it like the AI that they are usually scared of? And how do you overcome this change in digital transformation? Yeah, I mean, I think we're all, I think especially the older we get, the more we become laggards, right?
00:44:57
Like I was quite proud of myself. I tried Huel for the first time the other day, you know, that kind of like food replacement where you mix it as a powder. Yeah, Huel, I tried that for the first time the other day and I was kind of always a bit kind of like skeptic, too dystopian for my liking. But I actually decided to go for it. But I think that's, I probably would have had it a while ago if I was a bit younger. So I think the older we get, the more we kind of get a little bit more, we have more scars on us in terms of like, you know, being burnt by things in the past. And we've become, I guess, more selective in what we apply our time to.
00:45:33
I think that's what we see in digital transformation is that often people are stretched. Taking on a new project is quite a lot of work. And you have to really make sure that you are able to make sure that it's worth it, right? There is a good ROI attached to that to make it worth the pain, you know. And I think that is kind of one of the things that we see with digital transformation is that people are quite skeptical to get started. But once they get moving and once you show early results, you can win a lot of people over. I can't remember your question anymore. But I'm just yabbering on about digital transformation now. But in my opinion, yeah, I think that's kind of, it is difficult to win people over.
00:46:24
I think what we see, especially within operations, because it is particularly an area which they don't want risk. They're really busy doing their current jobs and firefighting. It's quite hard to drag them out of that mindset to do something different. And I find what really works well within that area is to come from their world, see things through their eyes. And also have the empathy to walk down onto the production line and understand. And the conversation changes a lot from being resistant to actually being acceptance. What we found, and this is an approach that we took at Adidas, was you can try top-down. Top-down works, of course. You can try bottom-up. That's also a long journey. But the two of them together work very well.
00:47:14
If you're able to gather insights from the ground level and communicate them up and bring that back down and vice versa, you can achieve a lot. And if you try to tell someone this is the right way to go, they'll often do the opposite and try to do it a different way. If you work with someone and say, what is your way? How would you like to use this technology? What insights would you like to have? That baby becomes their baby. And they start to then grow that child themselves. And that digital transformation then happens in a more organic fashion. So I always find that it really works when you actually go and be empathetic to the people that are going to use the technology. And that then adoption comes very naturally from there.
00:47:55
Again, I can't remember your question, but I've answered, I think, in 12 different ways. Sorry. I was a bit intrigued indeed by, let's say, as we age, we become a bit like ourselves because of the scars that we gather along the way. I got intrigued by it because I'm a profound, let's say, initiator. We need to embrace change and things like that. But I also feel sometimes myself like it's less when I was a bit younger than I am now. It's not like I'm old, but I still feel like my tendency to embrace change, it's more tricky than it used to be. And I think sometimes, Marie, you also feel like when we have discussions like, oh, maybe we should be doing that. Like, let's wait and see before we change things.
00:48:47
Whereas like in the past, I think I would have just jumped on the opportunity to change things. Which is not a bad thing either. Like you shouldn't try to do everything, which we sometimes try to do when you're younger. It is a balance, you know. There are some things that you should challenge and some things you shouldn't do. And there are some things that you should evaluate first and go with. And again, that kind of goes back to that cultural approach that, you know, we compared China to, say, Europe. And not saying either methodology is wrong, but as long as the end result is the right one. What you got there in the end, you know. Definitely. I'll answer the last question from my side as well.
00:49:27
If there would be anything that you could change at the start, let's say, of your journey. Would there be something that you would have changed yourself? A hundred percent. A lot of stuff. You can only change one thing, Scott. I'm no angel. I think something that I've kind of realized over the past few years. And we did a course during the MBA called Power, which is a very kind of strange title for a course. But it was really interesting. It was essentially understanding how to build up networks, how to get yourself out of your comfort zone, to lose sometimes your inhibition and be able to then go and speak to people. And from not just doing that course, but also kind of practicing it in the roles that I play in the past few years.
00:50:21
I've realized that the world is a very open place and people are very open to talk. and you can essentially speak to anybody you want and sometimes, which is fun, and I think that is something I wish I'd done more when I was younger is kind of not got so stuck up and I shouldn't be allowed to speak to these people or I shouldn't be able to communicate with these people and the world is quite an open source ecosystem where you can speak to sometimes whoever you want if you have the right reason to and I think that's a great thing for if I was to tell myself when I was younger is be more bold and don't be afraid to go talk to people and collaborate and network etc.
00:51:06
I think it's a necessary talent especially in digital transformation where you're trying to win over stakeholders, you're trying to get resources allocated to your projects, you're trying to do these things and that skill set enables you to achieve and the organization to achieve more which I think is a super important thing to do. Be proud of what you're doing you know. That's definitely true and I think it's a really great final call to action from the podcast towards our audience like just go out there and start talking to people because like people love to talk about their own experiences and you're gonna learn so much from it. Just to wrap it up I think we learned a lot about AI, artificial intelligence in the manufacturing setting where it's really leveraged to go from a more reactive state to a proactive state. Thank you Scott for your insights and yeah we'll be with you very soon with other insightful guests but before we go there again Scott thank you so much for your time and looking forward to see you again in London man like next time we visit we'll definitely send you a message. Yeah drop in, we'll catch up for a beer. Cheers, thanks very much guys it was a pleasure. Thank you.
