S1:Ep74 Amazon’s Blueprint for Success: Elevating Company Culture in the AI Era

In this episode of Digital Transformation & AI for Humans, I’m joined by my guest from Luxembourg – Attila Lengyel.

Attila is a former Amazon Web Services senior leader, Amazon Bar Raiser, Chief Innovation Officer at EduGamiTec, keynote speaker, strategy facilitator, innovation coach, and company culture expert.

By leveraging Amazon’s proven mechanisms for innovation and customer obsession, Attila has helped companies like Securitas and HEINEKEN create digital products and services rooted in real customer needs. As an Amazon Bar Raiser, he has conducted over 500 interviews, supported over 100 hires, and mentored future leaders. His career also spans more than a decade at Microsoft and Exact Software, where he translated business objectives into transformative technology solutions.

🚀 Key themes we explore in this episode:

• How AI amplifies Amazon’s culture of customer obsession and operational excellence
• Leadership strategies for scaling AI with agility, accountability, and ethics
• Why AI should enhance human creativity and decision-making – not replace it
• The mechanisms Amazon used to integrate AI without disrupting its core values
• How companies can future-proof teams through reskilling and continuous learning
• AI-driven personalization: examples of hyper-personalization that build customer loyalty
• Preparing for the AI singularity – what leaders must do now to stay relevant
• Attila’s most important advice for leaders building AI-powered yet human-centered organizations

✨ Whether you’re a business leader, innovator, or culture builder, this episode reveals Amazon’s blueprint for sustaining growth, innovation, and culture in the AI era – and how you can apply these lessons to your own organization.

🔗 Connect with Attila on LinkedIn: https://www.linkedin.com/in/alengyel/

🌏 Learn more about EduGamiTec: https://edugamitec.com/

Chapters

• 0:00 Introduction to Digital Transformation and AI
• 2:46 Attila's Journey from Teacher to AWS
• 8:07 Amazon's Approach to Gen AI Integration
• 14:02 Elevating Human Performance with AI
• 21:32 Leadership Principles Guiding AI Adoption
• 31:16 Future-Proofing Teams in the AI Era
• 37:36 AI-Driven Personalization Examples
• 48:07 Preparing for AI Singularity
• 1:03:04 Final Thoughts on Human-Centered AI Leadership

Transcript

Speaker 1: 0:00

Hello and welcome to Digital Transformation and AI for Humans with your host, amy. In this podcast, we delve into how technology intersects with leadership, innovation and, most importantly, the human spirit. Each episode features visionary leaders who understand that at the heart of success is the human touch nurturing a willing mindset, fostering emotional intelligence and building resilient teams. In today's episode, we are going to talk about Amazon's blueprint for success. I am excited to discuss what it takes to elevate company culture in the AI era with my fantastic guest from Luxembourg, attila Lengel.

Speaker 1: 0:41

Attila is a former Amazon Web Services Senior Leader, amazon Bar Racer, chief Innovation Officer at Edu Game, tech Keynote Speaker, expert, strategy Building Facilitator, innovation Coach and Company Culture Nerd. By leveraging Amazon's mechanisms for innovation, atilla helped companies like Securitas and Heineken build new digital products and services rooted in customer needs. As an Amazon bar racer, he had contacted interviews, supported hires and mentored bar racer candidates. Attila's career journey also includes over a decade of senior roles at Microsoft and Exact Software, where he translated business objectives into transformative technology solutions. Welcome, attila, it's a great pleasure to have you here in the studio today. How are you?

Speaker 2: 1:38

I feel amazing. Thank you very much. It is an absolute pleasure to be here in the studio. Thank you.

Speaker 1: 1:44

Fantastic. Let's start the conversation and transform not just our technologies but our ways of thinking and leading. If you are interested in connecting or collaborating, find more information in the description below. Subscribe and stay tuned for more episodes. I'd also love to invite you to get your copy of AI Leadership Compass Unlocking Business Growth and Innovation the Definitive Guide for Leaders and Business Owners to Adapt and Thrive in the Age of AI and Digital Transformation. Find the Amazon link in the description below. Attila, to start with, I'd love to hear more about your journey, about yourself, about your professional achievements. Could you tell us more?

Speaker 2: 2:46

teacher of history and economics. But back in the 90s I'm originally from Hungary and back in the 90s being a teacher in Hungary was not really giving you super cool career prospects. So, as Hollywood stars say now, I was young, I needed the money, so I just moved on to IT Never on the technology part of IT. So I didn't become a software developer which many of my fellow students moved on to. I always stayed on the business side. I always focused on understanding business objectives and then use technology to make that happen. I think from the very beginning I was focusing on how to allow companies to better themselves, how to enhance themselves themselves, how to enhance themselves. At the beginning it was very easy because it was mainly about a certain function like warehouse or customer service or internal procedures, and I was showing a piece of software which I told customers that hey, buy this and then you will be better in this or that aspect. And then, about 11 years ago, I've been approached by Amazon Web Services and they offered me a possibility, an opportunity to work at AWS and start working with customers and start selling cloud computing, which was actually a massive step up, because cloud computing and probably we're going to touch this later on a little bit. Cloud computing was an absolute new way of doing IT. So suddenly, instead of just focusing on a single function, I was having conversations with companies about transforming their whole operation, or at least certain aspects of their operation. It was very exciting times, super exciting times. I kind of observed 2014, 15, 16, when this new paradigm of IT transformed from being a nice-to-have thing which is only good for garage developers and it just transformed into this enterprise new way of doing IT. Very exciting times.

Speaker 2: 4:36

I was doing it for about two and a half years and then I moved over inside AWS to a new team, the training and certification team. Because what I kept hearing from my customers is that, attila, we get it. We see the business value, we see why would it make sense for us to move our data centers and just leave them behind and move to the cloud? But we don't have people. We have people who have decades of super important expertise. They have no idea about the cloud. So I thought, okay, that's going to be the next big thing. I moved over to training and certification and started to build large-scale internal education programs for enterprises and trained thousands and thousands of IT professionals to transform their knowledge into the cloud.

Speaker 2: 5:24

And about four and a half five years ago, I started to hear new things from the market when they told me customers told me that, yes, we get it, and now we also have the people who can actually do this. And we did the basic lift and shift. We kind of did the basics. Now what? What is it? What's the big thing? How is this new IT going to really transform and change things in our company? So then I thought, okay, so what's next? And it was innovation. So what this new technology enabled is to be super fast in coming up with new ideas and make them happen.

Speaker 2: 6:05

So I joined another team inside AWS called the Digital Innovation Team, and what I did in the past four years before I left AWS was working with companies, telling them the story how Amazon innovates and I'm pretty sure we're going to touch this topic later on in this interview how Amazon innovates. And if somebody said, okay, that makes a lot of sense, it's. Can we give it of sense? Can we give it a go? Can we think like Amazon? I was working, I was running workshops with them where I coached them how to get from a very specific customer problem and how to work through it and think like Amazon in order to get to a very targeted and specific new product or service. So that's what I've been doing. And then I decided it's time for another change in my life and I made a big leap of faith and joined a startup who has a very revolutionary idea about how corporate training should be done in the future. So that's EduGami Tech, and that's where I'm responsible for business development and innovation at the moment.

Speaker 1: 7:04

Such an exciting story Sounds like an amazing roller coaster and so many new learning curves.

Speaker 2: 7:13

Yes, that is true, that's absolutely true.

Speaker 1: 7:18

Super impressive and that's how we grow. There is no other way to grow and develop and learn something new in life.

Speaker 2: 7:28

And I think I can be very grateful that Amazon has a few really good mechanisms which enables you to grow, so I think I can be grateful for that. It was fun.

Speaker 1: 7:41

Sounds great. Amazon thrives on customer obsession and operational excellence, besides supporting those new learning curves and talents. So how has artificial intelligence amplified this and what strategic shifts have been most effective in keeping Amazon at the forefront?

Speaker 2: 8:04

shifts have been most effective in keeping Amazon at the forefront. Yeah, you can imagine that, since the time that GPT popped up and became common knowledge when was it like? November 2021? Since that time, gen EI was everywhere inside Amazon, inside AWS, but pretty much every single company I've worked with in the past three and a half years, they were all focusing on this new technology and tried to figure it out, how to use it and what to use it for. So the fine print here is that I was not responsible for deploying Gen AI inside Amazon, right, but I had the privilege and the love I was collaborating with a few of those people who were working on the Amazon retail side and were responsible for leveraging this new technology, or those who were also developing on the AWS side, the latest and greatest Gen AI tools. So I had some possibility to take a peek behind the curtain and see what's going on behind the scenes, and I also had my personal experience based on how AWS and Amazon was leveraging this new technology.

Speaker 2: 9:10

Based on all this, I can tell you that I observed that things were going two speeds. There was a slower and a much faster swim lane. I think it was. First of all, it was mind-bogglingly amazing how fast Amazon made this new technology available for their internal employees. So that's the fast track. They made it sure that the moment there were anything tangible, it was already opened up a sandbox environment for every single Amazon employee and everybody was strongly encouraged to go there and just start playing with it and also created a lot of educational content. So even amateurs like myself, who has zero developer background, I was able to start just playing with it. I think the mental model was that you have to understand what it is if you want to sell it, like in AWS, but you also have to understand what it is if you want to come up with ideas how to leverage it in your everyday life. So that was very fast.

Speaker 2: 10:12

On the other hand, what was really slow is the speed of creating new customer-facing products and deploying them in real life, especially compared with a lot of other companies I've been working with who were just breaking their neck in order to push out the first Gen-EI powered whatever into the market so that they can be the first. Amazon was not really bothering with that. It took quite a long time, for example, to get the very first real application of Gen-EI in the Amazoncom retail websites. I think it was in August 2023, so almost like two years after the whole Gen-EI craziness started, when you started to see the first tiny little fields on a product page which said here's what customers say about this product. These are the top three characteristics they call out and they defined print saying this was created using a Gen AI tool. It happened in August 2023.

Speaker 2: 11:16

And then the real big sales assistant Gen AI powered sales assistant called Amazon Rufus was launched in 2024. Called Amazon Rufus was launched in 2024. So like in the middle of 2024 and just in the US. So it took Amazon like three years to really figure it out. How can these Gen-AI technology be used to delight customers? So that was my observation that the customer obsession manifested in this way that, instead of just rushing something to the market just for the sake of being first using a technology, they really thought it through that. Okay, so keeping the customer focus, how can we actually get something meaningful out of this?

Speaker 1: 12:01

Such an interesting development and I've been following on that as well. And when Rufus got introduced, within a week I had an interview with somebody who has been on Amazon's side and it was really impressive. So interesting to see how everything is developing. But sometimes I agree, it's better to take your time and put everything into place to expand from there in a sustainable way.

Speaker 2: 12:38

Yeah, absolutely. Amazon culture really manifests in everyday work is that what's being said to every single employee inside Amazon is never fall in love with the solution, never fall in love with the technology Although, trust me, sometimes it can be super hard because, especially if you have a technical acumen, you will feel like you are in a candy store, have all these super fancy tech texts just at your fingertips. But they say that never fall in love with a technology or a solution. Always fall in love with a problem, a customer problem, whether if it's an external customer or an internal customer. Focus on that, fall in love with the problem and then just find the best solution. Even if it's a 20-year-old technology, it doesn't matter.

Speaker 1: 13:26

I love this approach. That's so wise and it is really something what helps prioritizing what really matters and solving that problem, Because I agree it's so easy to fall in love with the solution Absolutely easy to fall in love with the solution. Absolutely, Attila. How can leaders ensure that AI enhances decision-making and innovation without eroding human creativity and autonomy? Can you share a success story where AI strengthened the human role rather than replacing it?

Speaker 2: 14:02

Yes, that's a very relevant question right now, but I've already mentioned that I'm a history teacher, right? So, honestly speaking, I've seen a few occasions where a leap in technology, in technology development, had an impact on society by, for example, making professions obsolete, pushing people out of their profession, like. Think about the industrial revolution, right, when industrial metalworking became a widespread thing and then suddenly you could see blacksmiths being pushed out of their job Because, although they were creating very delicate and high quality horseshoes and tools and whatnot, it was just much easier and cheaper to buy the ones which were created in massive factories. And the same happened in some other areas. So I know that right now, gen AI is having a very similar impact, and I also know that it's not fair to compare to the industrial revolution, because what's happening with Gen AI now is, I don't know, like a million times faster, and usually technological revolutions in the past impacted a very narrow group of people, like certain professions. Now Gen AI is impacting hundreds and hundreds of professions. So that's why it's a different, slightly different setup, but what I have seen so far is these tools can actually elevate human performance, very similar to what happened with the graphics artists back in the 90s when Photoshop became a thing. And then I heard because my wife is coming from that industry, so I have friends and ex-colleagues who can tell us that back in the 90s I have friends and ex-colleagues who can tell us that back in the 90s there was a very strong fear in the graphics artist and the graphics professional community that Photoshop will allow every single stupid amateur to do beautiful art at home using their PCs and they will be not needed anymore. Now, after a few decades, we see that that's not what happened. What actually happened is that those very simple tasks which previously required a professional, because the equipment or the things, the knowledge is needed, because it was not available for everybody, those tasks suddenly became available for simple home users like myself. So my mom had to go to a photo shop to create invites for my birthday party right Back in the I don't even dare when it was, but pretty much in the 80s. And now, when I did it for my own kids, I was doing at home, right. So these simple tasks which don't really require high-level intuition or artistic view, this has been taken over by these technological advancements. But the people who, actually those professionals who embraced this new tool set and added it to their own toolbox, they actually were able to elevate their artistic impression into a much, much higher level, or they were able to do same quality work but in much higher quantities with much higher speed. So I think something similar is going to happen, and I have seen the first signs of it.

Speaker 2: 17:13

We were talking about Amazon culture, amazon, if I look at how Genii is being used inside Amazon and just recently I talked with one of my ex-colleagues and he confirmed that what's happening right now is this Amazon has this very specific narrative culture. I've probably already heard about this that we didn't use PowerPoints inside the company when it came to certain meeting types or certain conversations. You are not allowed to use PowerPoint. You have to write complete documents one-pager, three-pager, six-pager documents with full sentences, full paragraphs, and the logic behind that is two things. First of all, if you are reading a document, some psychologists calculated that you can digest 150% more information than watching the same thing on a slide. Okay, that's one. The second, when you are writing the document, it forces you to really think it through, because throwing a few bullet points on the PowerPoint, that's okay, but if you want to create full sentences and full paragraphs, you really need to understand what you want to say. So we have this culture and it is challenging. It is not easy, I can confirm. I don't know how many documents I wrote in 11 years, but a lot and it is very hard.

Speaker 2: 18:35

And now, for example, one of the things what they use Gen AI for is writing these documents, not taking over completely the whole document writing process, but to create the first version. You're probably familiar with this blank sheet paralysis, when you have to stop writing something and all you have is the blank sheet. And I am like that. I have this paralysis. Okay, how do I start? Where do I start? Having a Gen AI tool which is capable of generating Amazon style documents is a massive help in this. It can accelerate greatly the whole Viking process because you fit it up with all the data you want to use and want to leverage and it creates the first draft, and that it's much easier to iterate the first draft than to start it from scratch, at least to some people. Some people probably think differently. So that's one example when Gen AI actually accelerates or elevates human performance.

Speaker 2: 19:35

Yeah, and this was one of my projects with a Swedish customer. They have 100,000 patents, legal documents which they want to leverage when they are serving customers and all their salespeople is not just encouraged, but it's their job to listen to the incoming customer requests and they'll look into this massive wealth of innovation which they have on paper and figure out how can they create something unique, something better what everybody else on the market can do. But the challenge is that through acquisition, through their own innovation process there are like more than 100,000 of these and these are legal patent documents very hard to comprehend for a normal human like myself. So they felt that they are leaving a lot of business opportunities and a lot of knowledge dormant and just on the side and they are not using it anymore. So they were thinking, ok, how can they address this?

Speaker 2: 20:35

So we had this innovation workshop with them, workshop with them and eventually it came out that what they need is somebody who would be able to read through all these documents, digest their knowledge or the information stored in them, and then have a meaningful conversation by understanding customer requirements and then look through all these innovations and bring the most relevant forward and even make connections between the request store potentially a sales guy looking at the request for proposal document right now and the people who created the patent, so the original owners of the idea, and eventually it turned out that this should be a Gen AI tool, it should be a Gen AI agent which is capable of doing this, and the last time I talked with them, they were already thinking about deploying at scale internally this whole tool, because they proved to be super helpful. So that's what I say, that elevating human performance. There are many aspects where Gen AI tools can actually do that. Instead of pushing you out of your role, it's rather giving you some leverage to do your things better.

Speaker 1: 21:50

That sounds really good because that relevance, that closeness to the actionable insights, of course it creates a different level of development, of growth and it helps moving forward in a more powerful way. So that's a great example. Thank you so much for sharing.

Speaker 2: 22:12

But it's not an easy ride for the company Because in many cases it happens that if you want to leverage this new technology, you have to change things. You have to change the status quo, your behavior, you have to do stuff differently, and there are many cases there is a pushback from your employees inside the company. And I have seen similar situations 10 years ago, when cloud became an enterprise-ready thing and suddenly large enterprises and companies started to think about okay, actually, yeah, we should move to the cloud, we should leave our on-prem infrastructure behind. I had this customer in the Czech Republic, a media company, where the CTO I mean, he got it. He understood it from the very beginning. He was so much into the whole cloud computing concept that the first time when I met him I didn't have to do my sales speech. He was telling me exactly the same arguments that, yeah, he gets it and he wants to do that.

Speaker 2: 23:11

So he told me an anecdote that when he announced it to his team of 70-something IT professionals and announced him that, yes, we are moving to the cloud, and then there were a few people who became super concerned. One of his mid-level managers raised hands saying that hold on for a second, hold on. Right now, 50% of our time is spent on managing our on-premises infrastructure replacing hard drives, plugging cables or whatnot. If we move to the cloud, this will not be necessary. So this means that are you going to fire half of your people? And then what the CTO said was I think it was very cool what he did he reached into his drawer and take out an A4 paper with a very dense list up and he said gentlemen, this is the backlog we have based on all the requests coming from the business. Right, these are the things you were not able to do because you were exchanging hard drives and plugging cables. If they move to the cloud, you are going to work on this. So nobody's going to get fired. Actually, your job is going to be more exciting than it was before.

Speaker 2: 24:24

So I think that's a really good story to tell what can be expected and the follow-up of the story. So the aftermath of the story is also a sign of what you have to be prepared for as well, because a few of these people who he was working with in his team actually they were just resisting. They were more interested in keeping the status quo and just keeping stuff and do stuff as they used to do it, so that eventually they had to be pushed out from the organization Because they were just resisting this change. They didn't want to do that, they didn't want to go there. They were just bringing up arguments why not to do, why it won't work, and it became quite toxic for the whole organization. So they had to be pushed out. I can't help, but I had this analogy about this story in my head and I can see something similar happening with Gen AI as well.

Speaker 1: 25:12

It is very similar, I agree. However, I just want to invite ourselves and all of our listeners and viewers to keep in mind that this is something super powerful, super impactful, and there is a risk that it is going to go in a direction we don't expect right now. So we have to be a little bit more attentive and just keep in mind that the historic data is not always something we have to refer to, because this is something quite new. So, yes, I totally agree that that example is perfect and in so many cases, it enables us as professionals, as humans. It enables us as professionals, as humans.

Speaker 1: 26:04

I'm guilty myself. You know I love using AI in so many different applications and I find it incredible. I'm in love with it. However, I also see how sometimes it starts overriding the human centricity in the organizations, and that is probably not where we would like to come. So leadership is important. Let's keep talking about leadership a little bit more, and I wonder what leadership strategies are essential for scaling AI while maintaining agility, accountability and ethical responsibility, and how do Amazon's leadership principles guide AI adoption in a way that ensures sustainable success?

Speaker 2: 26:52

Yeah, I think it's very important to see. When I was talking about that super fast swim lane of making AI, of Gen AI, available for the employees, the reason why it was working and the reason why people were not really abusing it but just really using it as it is intent to, was very much rooted in the company culture of Amazon. So the reason why it is absolutely natural to open up these new experimentations, these new tools, to employees is because everybody has an understanding about the common goal of the organization, the direction that they are going and the priorities and the values, what the company is looking at, and these are captured in the leadership principles right, so there are 16 leadership principles. Leadership principles right, so there are 16 leadership principles. And I think how employees approaching this new technology is a very good example of how these principles work in real life, because if you look at these LPs, they are public. You can read all of them, read what they do. But it's very important to understand that these are not black and white decision-making matrices and you cannot excel in all of them at the same time. That's meant to be like that. These are guideposts, so this means that you, as an employee, continuously have to balance out or create a balance between applying these leadership principles. So, for example, if you are at the beginning of the project, then probably diving deep is your leading principle, because you want to understand what the data you have and what's in the background of certain things. And as you learn more and more, at one point, bias for action is becoming the leading principle and diving deep is losing a little bit from its importance because you already know enough. It's now time to make a decision, and I think there are multiple leadership principles at SideAmazon.

Speaker 2: 28:47

Which has an impact or is guiding people when working with these new technologies? Number one customer obsession, right. So I think that is the top. That's the absolute leading leadership principle, and every time when anybody looks at any type of new tool, new technology inside Amazon, that's the first thing we ask how is it going to help our customers? And if there is no clear answer, nobody's going to push it. We don't really care how it's going to help the customers, because we need to have a Gen AI tool out in the market. It's never going to happen. On the other hand, invent and simplify is just as important. So this means that we really need to look into customer need and work backwards from there.

Speaker 2: 29:31

But also there is frugality, which to me, it was a weird leadership principle at the beginning and I think I misunderstood it for quite a long time. I thought in the first one or two years that frugality means being cheap, like saving money anywhere I could, until one of the leaders, one of the senior leaders, explained to me that no, frugality means achieving more with less, so which means that it probably shouldn't be your leading principle, but you have to have it in the back of your head Every time you are about to make a choice. You have to question yourself Okay, first of all, what else can I use this thing for? Or is there anything else I could use instead of building something new, so kind of trying to stay frugal with your resources, and so on and so forth. So there is a lot of it.

Speaker 2: 30:22

And earning trust, earning trust of your colleagues, earning trust of your customers that's probably even the second most important of the customer obsession, meaning, the moment anybody inside Amazon is dealing with data whether if it's internal data or customer data they have to be absolutely sure that they comply with all the data regulations internal and external data regulations Every time they are trying to use, for example, a Genial tool.

Speaker 2: 30:52

They have to make sure which data is being used, how is it being used, and no risk is going to be opened up to our everyday operation by using this tool.

Speaker 2: 31:04

And the thing is that these are deeply embedded into everybody who spent a few years with Amazon. So this means that of course, there are a lot of regulations and governance inside Amazon. There's a lot of things which just comes natural. You're never going to open up any data source to the public or to make it comfortable to use a new gen AIT right, because you want to earn trust and things similar like that. So when you have all these guardrails in place, then actually you as a company leader can just simply open up these new tools for your employees and just get out of their way and just let them experiment, let them explore, let them figure it out how things can be used and allow them to think freely, and eventually they will come up with brilliant ideas. There are probably hundreds of thousands of new ideas popping up inside of 1.5 million employees of Amazon, because they know that they can bring these ideas forward and there was one additional cherry on the top.

Speaker 2: 32:22

Yeah, this cherry on the top was that this mentality, a customer obsession and earning trust and diving deep. Actually, this culture brings technical and non-technical people at the same table. This is I've been observing in all the customer engagements I had in the last four years. One of our prerequisites was when we were doing an innovation workshop is to bring the business people and the IT people to the same table, and that's what's happening with Gen AI as well, because it's a technology advancement. Right, it's a technological tool, but the use case is going to come from business, from everyday life. So making it possible to experiment with them naturally will bring people who have common interests but probably very different skill sets and priorities, and bring them to the same table. Bring them to the same table.

Speaker 2: 33:23

I don't know how many times I heard from CEOs of companies telling me that after a Working Backwards workshop, they said these people, like the product teams and the IT teams, never, ever talked to each other in the past. This was the first time when they actually started to think together and they realized that, in the end of the day, they have the common overall goal, that in the end of the day, they have the common overall goal and they were able to start building compromises, giving up things to make things happen. So that's the cherry on the top. If you are doing it right, then your business and your technology people will start having conversations with each other.

Speaker 1: 33:58

That's actually my favorite part and you know it warms up my heart that you mentioned it, because that's exactly the gap many companies are suffering from and it is still there, that silo effect. It is still there. It is still reflecting on the overall results. It's not that difficult to deal with it, but it starts with the mindset. It starts with understanding its importance and also with the plan how to approach it, how to put the people together and make them understand each other's pain points and goals.

Speaker 2: 34:41

Exactly. And they are not going to do it by themselves. Because unless you have a very strong company culture which, yeah, I've seen a few companies but in most of the cases company culture is just a marketing fluff which looks good on your website but in the end of the day, nobody really has a common view of what your company culture is, and in that case, everybody inside the organization will have their own idea about what should be important and how they should be making decisions, and that will create conflicts, right? Because the head of IT and the financial leader and the marketing leader will have completely opposing interests if they are left unchecked, and that will build silos, that will cut communications and eventually it's going to hurt your company.

Speaker 2: 35:35

And you know it's very much two-way. So don't misunderstand me. I'm not saying that, oh, those poor IT people, they're not, nobody's listening to them. No, that's not what I'm saying, because I also have people around me who are high-level IT leaders, people leading IT organizations, right, and I can hear when they're talking about the business, the other units, that oh my God, they are so dumb?

Speaker 2: 36:00

Oh my God, they don't understand it. And if somebody comes up with an idea from the business side, for example, the first reaction is oh, that's never going to work for this or that or that reason. So I'm saying that this silo, this zero communication, that's not trusting each other. That's a very much multi-directional, the common, common, common thing.

Speaker 1: 36:22

And you know, I recognize those trends and I believe you have to be one of them and belong to the tech side of things as well in order to have access to those conversations. But when you are there, you can discover a lot and that's how you can also find the ways to introduce more communication and those common initiatives, common conversations.

Speaker 2: 36:55

Yeah, absolutely.

Speaker 1: 36:58

Let's talk more about the culture. Many companies struggle to scale AI without disrupting their culture. What mechanisms did Amazon use to integrate AI successfully while staying true to its core values? What would you recommend to leaders who want to preserve and evolve their company culture in the AI era?

Speaker 2: 37:19

company culture in the AI era. Yeah, that is a very valid and relevant topic, because these new tools can be abused right and in that case it will create something very toxic and very harmful inside the organization. So I think the only way to avoid it is if you have a strong culture, if you have a strong common sense, a strong common goal which every single people inside the organization understands and they know that. Yes, this is the purpose of our company. These are the type of customers we want to serve with these type of services. These are the values we have, and these are not just fancy words written behind the receptionist desk, but these are being encouraged and enforced on a daily basis. You know, it's like when a universal tool in the past was introduced like Microsoft Excel.

Speaker 2: 38:16

When Microsoft Excel became a thing and it started to appear first oh my God, I remember it was Lotus 1, 2, 3, when I was doing my exam from spreadsheet softwares. So I know that at that time Excel was first. It was used by a very specific group of people for very specific purposes. There were the ones who are working with data and they were doing simulations and so using like 70% of Excel features, and then there were people who were just using it to create some fancy cables and make some nice graphs. If you look around now, excel is everywhere. On every single laptop and machine you have Excel and you're using it on a daily basis. So it became a tool which became absolute universal and absolute widespread.

Speaker 2: 39:04

And the thing is that at the beginning it was actually a concern. How will people potentially abuse this tool and create fake data and create things which look brilliant but actually there's nothing behind them? Because the tool made it possible and I think the same is happening with Gen AI and it's like exponentially higher level, because if you think about the document writing example I told you, if it's inside Amazon, because they have this leadership principle to insist on the highest standard and earn trust. Therefore, I see less of a risk that somebody just overuse and abuses Gen AI tools and technically just leans back and let whatever tool they have to write fancy documents and not putting any effort into the whole process. But if you lack of that very strong cultural guidance, then yes, there is a risk. Your people will abuse these tools and they will create good enough results and get away with it and eventually that might hurt your company, because your intellectual power is just going to diminish because everybody will be happy to create that document. It looks good, a lot of fancy words, off you go and then just do that thing.

Speaker 2: 40:29

So I think that's why it's very important to have the company culture. That's the only thing that might well, that's one of the things that might prevent this situation. And I think culture is very important because it cannot be replaced with governance. I mean, I've seen companies trying it, trying to limit the use of ChatGPT or Microsoft Copilot, and they are putting some super smart, sophisticated tools which can detect if a document was created by a Gen AI tool. These can be tricked all the time and it's just going to be a never-ending vicious circle how your employees are trying to trick whatever governance you put in place, if they don't have the intrinsic motivation to do the right thing, which is coming from company culture.

Speaker 1: 41:21

Exactly, and it's also about talent retention. It is about human perspective, that you want to enjoy your time during the working hours as well and you want to see people giving their best to your business, to your company, and that is possible only if they want to. It cannot be really forced, because it can be forced until a certain point and when that limit is reached, you can't push more than that. It will just get such a resistance that it will stop there, and that's really important to keep in mind, I think.

Speaker 2: 42:07

Exactly.

Speaker 1: 42:09

Let's take a look into the future, Attila. As AI reshapes workforce dynamics. How can companies future-proof their teams and foster a culture of continuous learning? What lessons can we take from how Amazon upskilled or redeployed talent in response to AI, and what does long-term success look like in an AI-augmented workplace where people and technology work in synergy?

Speaker 2: 42:42

That's the billion-dollar question, right, and a lot of companies are trying to figure this out how this works. What I have seen again what I have seen is Amazon is really good in encouraging people to push themselves a little bit over the limit, go over their job description, and I think these Jennyite tools and opening them up to their employees was a very good example for them. Because here's what happened these sandboxes were opened up and everybody was highly encouraged to go there and dive deep and start educating yourself, but, of course, we were not at the same level of interest, in the same level of being invested into this idea. Some people were more interested in it, some people were less. What actually happened is that, regardless of your job role and your background, there were people who actually started to become subject matter experts of these Gen AI tools, even if it was not their job. One of my very kind colleagues, amir Ilyan, who was a colleague at that time, that's where he really started to dive deep into all these Gen AI tools and after a year, he became our go-to person in our team when it came to any type of Gen EI related questions.

Speaker 2: 44:05

So I think, if you are talking about, okay, what could be. How can we make it future-proof? How can these dynamics be influenced a little bit to make it helpful for your organization? I think this is it. This could be a possibility to just pollinate your organization with this knowledge. Just let subject matter experts pop up across your organization.

Speaker 2: 44:32

Do not limit the learning and the diving deep into this technology to certain job roles and keep them in silos saying that, no, it's just for data scientists who are sitting in IT. But make these tools available. Let people experiment, let people play with it Of course, it's certain rules and governance in place and allow people to get to the level of knowledge which they feel comfortable with people to get to the level of knowledge which they feel comfortable with. So I think, because Gen EI is going to be present in every single aspect of our work, you need to have these little hubs of knowledge being present everywhere in your organization. If you think about access, in every single team you know who is the person who can create the pivot table or who can create some very sophisticated Excel feature, because that's her thing, that's her hobby. I think something similar will have to happen and will happen with the Gen AI tools as well.

Speaker 2: 45:38

I love pivot tables and how there is that one person yes there is, and you know that somebody will come to you and say, hey, you did that thing and it was so cool. Now I have this problem. Can you help me? I think it can be solved with Excel. Can you help me? Exactly the same conversations are happening with Genii. I agree, but again, it's a cultural thing, so that your people need to have the motivation to develop themselves. You need to have an environment when people are not punished if they spend time on making themselves better in certain areas. And yeah, you also have to be okay with the costs of these things, because if you open up these tools and people will start experimenting with it, they will come up with ideas, they will start building new projects around it, at the beginning, just internally, later on maybe for customers, and there will be some failures, right, and you as a manager, as a leader, you have to be okay with that and not push out somebody from the organization immediately if one of their projects fail. So again, culture.

Speaker 1: 46:55

True, and that's the problem, the typical problem about innovation and those pioneers within the organization who are propagating new approaches, new ideas. And of course, it takes resources. It requires time, money and effort and it's not always welcome.

Speaker 2: 47:17

And it's not always welcome. Just one last thought on that. Actually, inside Amazon it can be a very good value for you to progress in your career. So to get to promotion, becoming a subject matter expert of a technology, of a solution and offering your knowledge to your colleagues, participate in projects and bringing your knowledge to the table. It is highly encouraged by even the career building or career progression process.

Speaker 1: 47:45

Sounds great and it is so needed, truly so. Speaking about personalization, let's talk about AI-driven personalization. I wonder if you can share an example where AI significantly enhanced customer experience while straightening brand loyalty, and what future innovations do you foresee hyper-personalization and predictive customer engagement?

Speaker 2: 48:14

Yeah, this is one of the low-hanging fruits of the whole Gen AI movement. That was probably one of the first areas where Amazon also started to experiment, because personalization you're required to digest immense amounts of data of your behavior, of your choices, and then you set a machine learning model to predict right and Amazon has a long, long history Actually this is one of their core business models to provide this personalization and to understand your taste, understand your shopping behavior. So I think they are using machine learning. I heard that the very first machine learning model was deployed in 2007 at the large-scale production. That was for forecasting customer behavior. So it's not a single person's personalization, but forecasting the overall business. But just a few years later, machine learning appeared in the website recommendations.

Speaker 2: 49:10

Now, the same people who told me this, they also told me that there are two things which are usually the concern when it comes to judging the quality of the personalization. The first thing and what they need to keep in mind when creating new tools or when they want to deploy Gen AI tools like Amazon Rufus into this whole mix. The one thing is that customers, in many cases, they question if the personalization is real, because sometimes they have the feeling that just some dumb stuff is being put in front of them which has absolutely they have no idea where this is coming from and, honestly, sometimes Amazon is guilty with that as well. So if I buy a garden grill, why on earth does it show me another three garden grills? I'm not going to buy more of those, but it still happens. So that's one thing. So is this personalization really based on your shopping behavior or taste?

Speaker 2: 50:08

The second thing is questioning the authenticity of the personalization or the recommendation. Is this really what the engine throws out? Is this really what the engine throws out, or is it being biased by somebody's sales goal inside the organization, right? So, for example, when it comes to banks, financial companies, yeah, it's kind of it's the little voice in the back of your head that you use the chatbot or whatever, and you tell some data about your situation and somehow the chatbot will recommend you exactly the same for that again and again and again, regardless of what you type in. So I think these are two very important aspects which, in the future, needs to be kept in mind and I think what the Gen AI tools will have to focus in addressing these type of concerns and as far as I've seen so far. For example, amazon Rufus is actually pretty good in addressing these problems, and you were asking example of good tools. Try this one, Give it a go. I think it's fairly relevant and it's getting better with every single iteration and every single interaction you have with it.

Speaker 2: 51:23

I can tell an example from my own company. Right now, I'm working at EduGami Tech, and EduGami Tech is educational technology. We are using game-based learning in corporate education and we are also using what we call nano-learning. And we are also using what we call nano-learning, which is, technically, we break down larger curriculums in very small chunks just a few sentences, 30 seconds of video or 30 seconds of audio and we channel it through certain games like classical bijou world, whatever, non-contextual games. Right, these are not serious games, these are casual games, what you love to play, and then you have these tiny chunks of knowledge poking up during your gameplay.

Speaker 2: 52:04

For example, the bid you earned, you run out of lives or time or whatever. In order to continue, good, you need to answer these questions, and then questions come from our system and you answer it and, if you answer correctly, you got 30 more seconds to finish your level. So, and what we are working on right now is how to make these knowledge bits which are coming in front of the employees, being fully tailored to the situation, to the job role, to the seniority level, to the day in the week of the employee, and make it totally fitting and customized for that specific situation. And it is what is mind-boggling that this is possible Now with these tools, this kind of customization is just a reality, and we don't even have to have supercomputers somewhere in the background because it can be achieved. Okay, so that's another example for customization in knowledge. So you don't have to sit through a fixed curriculum, but actually you can get those bits of knowledge pieces which you need right now in your specific situation.

Speaker 1: 53:16

That's actually a great reminder. Your specific situation, that's actually a great reminder. I take it through my own prism and from my perspective, it also reminds me about the fact that human behavior is changing and our attention span is impacted. Now it's basically shorter than the golden fishes one. Yes, more or less. So we have to adjust our teaching processes, our methods of learning and sharing information with others so that it suits us as human beings and, instead of being overwhelming, it gets incorporated in our preferred activities and gets into our system. While we are really in that space of being open, being really available for that information and for those processes, we have to keep in mind that we're changing as human beings and our approaches need to change, and I love your example because it exactly reflects those shifts, and not only reflects them but also comes with a solution, and that's exactly what we want to see.

Speaker 2: 54:38

We are looking at it.

Speaker 1: 54:42

Attila. With AI evolving at an exponential rate, some experts predict we are heading toward singularity. How should leaders prepare their organizations for this potential future? What should companies do today to ensure they stay relevant in an era where AI might become the dominant decision maker?

Speaker 2: 55:07

Very pressing problem. I think it should be an ongoing conversation in every single boardroom and every single C-level meeting, because, yes, we need to prepare, and if somebody hasn't paid any attention to this so far, I'm quite afraid that they are already late, because things are changing exponentially, with an exponential speed. I would say that, yes, there are some absolute, mandatory steps which you as an organization, you as a company, have to take, and these are related to set security and ethical guidelines. So you have to create your own playbook and, of course, you can rely on experts from the outside right. There are people who are thinking about this for even decades, and you can leverage their knowledge and their understanding, but you have to have your own playbook, the rules which you're going to follow. The second thing is AI literacy. I think I already mentioned it quite a few times. Make these tools available for your people. Demystify Gen AI tools for your people, so that they understand what they are and what they are capable of, and do not think about them as some mystical, omnipotent something which they have to be afraid of. No, they have to learn that this is like Excel this is a tool. It can be used for certain things and cannot be used for others. So have this AI literacy.

Speaker 2: 56:27

I would say that you have to start building like a hybrid decision making mechanism, because I am pretty sure that with the speed of automation, we will get to the stage that more and more decisions will be made by Gen AI agents, and that's okay because that will speed up time. But you as an organization have to draw the line to what kind of decision have to or must have human control over. So there are decisions which can be fast and not. Yeah, which can be automated, but I think you have to create your own label, the hybrid decision-making. What else? There are like a million other things.

Speaker 2: 57:10

Oh, auditability. So if you are building your own Gen EI tools, I think this should be together with security, like job number one. I think this should be together with security like job number one. Make it possible to audit the whole thinking process of whatever model or whatever agent you are building, so that you will be able to trace back why certain decisions have been made, the way they were made. And I think that leads to, eventually, if you really start to build your own Gen AI large language model, so you start to build a fleet of your own tools. I think you should spend some time to tour in your company culture, your values, your decision-making process, your priorities and sort of frame, your own Gen AI agents which are in line with your company culture. So right now, these are the things I would absolutely focus in order to be prepared whatever is going to happen 5, 10, 20 years from now.

Speaker 1: 58:13

Great point and I couldn't agree more. Actually, explainability and our responsibility as leaders for avoiding those black boxes where we can't understand what actually happened and why it happened that way, it is something we have to make sure we keep under control, keep within our focus, because otherwise it might lead to the outcomes we wouldn't really like to deal with. But, attila, what do you think about singularity?

Speaker 2: 58:51

I think that it's getting closer and closer. You shared with me an article about GENI-AI models being able to replicate themselves and to my knowledge, that's one of the conditions of reaching singularity. So I think it's a really uncharted area for us. That's unprecedented. I talked about leaps in technology and new technology, advancements in technology and new technology advancements. None of them had this. I don't know this perspective, this potentially unlimited level of impact.

Speaker 2: 59:27

So this is new and I think we have to be very careful not to be arrogant, to believe that everything's under control, to believe that everything's under control and we are.

Speaker 2: 59:43

You know, we are holding everything in our hand, because it's very easy for things to slip out of our hand. And then, if we think about the speed, how, for example, hardware and compute power is also increasing with quantum chips and which is been announced by Google and also by Amazon, and I think all the other big vendors are coming up with their own very soon and now the compute power is, I don't know, like multiplied by a million. So suddenly these compute power will be able to power Gen AI models to do things at an unimaginable speed. We have to be really cautious to see what's happening and keep it under control. So that's what I think about singularity, that maybe I'm a little bit naive because I watched too many Star Trek in my youth and, as of still today. So I think reality will not be bad, I'm quite sure. But I really hope that we don't reach the matrix either, because that's also another one of my favorite, and I think hopefully the reality will be somewhere between these two extremes I love to extremes.

Speaker 1: 1:00:56

I love your conclusion. It is really good how you gave an overview. And, yes, those microchips and processors like Willow, for example, by Google. They are going to give access to completely different quantum computing power, which is going to take us into a space where we can't compete with the power of computers anymore in the same way as we do today having our usual solutions, and we're living in exciting times. It's both super interesting and frightening at the same time, and a lot depends on our decisions, on our choices and on our values as well.

Speaker 2: 1:01:47

Yeah, and I think, as a company leader, all of us has the power to guide where technology is going, because, in the end of the day, technology is mostly following the money right and the market needs and the company leaders who are supposed to be prepared for this. They are driving this need and I really hope that we will be able to make that trade-off to sacrifice some speed, some agility, some comfortable features in order to keep things under our control and not just giving everything, because this solution is 15% faster or 10% cheaper, but I have no idea what's happening in it, but I don't care, because I need that savings. I hope that we will be able to keep that balance 100%.

Speaker 1: 1:02:43

Thank you for pointing this out, because it's not only about performance, and sometimes those conversations lean towards performance and growth in its pure form, but there is always a price to pay and, as leaders, we have to keep it in mind. So, to wrap up today's amazing conversation I love it so much. I would love to continue asking you different questions and developing this deeper understanding, but what is the most important piece of advice you would give to leaders aiming to build AI-powered yet human-centered organizations?

Speaker 2: 1:03:25

Hmm, yeah, that's a good question, if I need one. It all points to the same thing, and that's your company culture. I can't bring up anything more important than that, and by company culture I really mean your values, your priorities, how you make decisions, how you communicate it with your people. So not the fluffy marketing words on your website, but really about the purpose of your company, because this is the only way to give an overall goal to all the employees you have to give an overall goal to all the employees you have, and this is the only way for them to understand how to use this technology in the right way, at the right place, at the right time, with the right purpose. Otherwise, I had no idea how you can keep under control any of these aspects of your operation. Yeah, that's it.

Speaker 2: 1:04:21

That's the best I can come up with.

Speaker 1: 1:04:24

That is truly the best and it is absolutely brilliant and spot on. Thank you so much for sharing your wisdom, your experience and your vision with us today. Attila, it's been a great pleasure having you.

Speaker 2: 1:04:38

Thank you very much for this opportunity. Really great conversation. Enjoy the second of it.

Speaker 1: 1:04:43

Me too, thank you. Thank you for joining us on Digital Transformation and AI for Humans. I am Amy and it was enriching to share this time with you. Remember, the core of any transformation lies in our human nature how we think, feel and connect with others. It is about enhancing our emotional intelligence, embracing a winning mindset and leading with empathy and insight. Subscribe and stay tuned for more episodes where we uncover the latest trends in digital business and explore the human side of technology and leadership. Explore the human side of technology and leadership Until next time. Keep nurturing your mind, fostering your connections and leading with heart.