Otto

What companies are leading in innovation in Europe in 2025?

Europe's Most Innovative Companies

The European edition of the Fortune 500, ranking companies by revenue.

Fortune 500 Europe

The first-ever European edition of the Fortune 500, ranking companies by revenue. 

F500 Europe 2023 Otto Group Bonprix Pilot Store

Otto Group ranks 216 with $17.3 billion in revenue (down 10.6%). The Germany-headquartered internet services and retailing company has 41,186 employees and made a loss of $320 million (down 115.4%).

Key Financials (last Fiscal Year)

Revenues ($M)

Profits ($M)

Assets ($M)

Total Stockholder Equity ($M)

Profit Ratios

Profit as % of Revenues

Profits as % of Assets

Profits as % of Stockholder Equity

Sector

Strongest dimension

Years on list

Country

Get insights into Otto including business profile, stock information, news, and rankings from Fortune.

Otto Company Profile, Stock Price, News, Rankings

legacy

otto

Companies

Otto | Fortune

Otto Company Profile, Stock Price, News, Rankings | Fortune

As generative AI encroaches on tasks traditionally handled by junior professionals—brief writing, memo drafting, spreadsheet modeling, legal discovery, and audit reviews—the apprenticeship model that has long underpinned law firms, consultancies, and accounting practices is under threat. AI replaces the bottom of the ladder, how do we train the next generation of partners, principals, and experts? Are firms ready to redesign the talent pipeline—or are they quietly bracing for a skills crisis a decade down the line?

Vanishing Rungs: AI, Apprenticeship, And The Future Of Professional Services

Conferences

The business world has already poured tens of billions of dollars into the AI revolution. But is this investment translating into meaningful gains in efficiency and productivity? And how are the most successful companies navigating challenges with adoption, including the high cost of building AI models, regulatory hurdles, and the impact on staff? We'll get a candid take on AI’s true potential for ROI—and how to deliver it in a way that works for consumers, employees, and the bottom line.

Inside the AI Revolution: Risk, Rewards and New Ways of Working

Mario Klingemann, Artist, Botto, In conversation with Tony Chambers, Studio TC & Friends

Brainstorm Design 2024: Androids and Their Connection to Electric Art

I think this one is gonna be fas fascinating 20 minutes with Mario where we are going to discuss Boto, the Decentralized Autonomous artist and the man who created Botto originally Mario. Before we get into the details on Botto, could you give us some insight and context to where you started? I believe you're a graphic designer working for nightclubs in Germany. Well, it's always difficult to kind of map my career because I always followed kind of what I was interested in most at a certain time and it was usually something that couldn't be taught. So yes, I was always fascinated by doing things with computers and code. And in the early nineties, there was a great opportunity because techno became really big in Germany where I'm from and they needed always kind of weird crazy graphics. So I became a graphic designer by writing Photoshop filters myself using algorithms. And this is when early, not really 1991 90 92 round. And uh so I would call it early generative art because in the end, yeah, you write an algorithm but then you add some sliders to it and get infinite variations and that's kind of my technique of, of doing my work and being kind of like the computer guy like that, which is also like, not really the label I want to have. But uh well as we all know, kind of the world evolved, the internet came along. So that's when those the same people said like, oh, there's a thing called website. So do you know what this is? So I made websites, then I made the video kind of like for German TV, on a computer, not on some million dollar machine. So it was always this end. But I always try to bring I'm lazy in that sense. So I love to spend like two days on an algorithm that then can do something for me a million times on the press of a space bar. So and yeah, so ultimately kind of that brought up the questions, what other things of myself can I replace with a machine? And of course, you said to me earlier that you always wished for an A I to do the things you didn't want to do even in those early days. So yeah, even like I think in 1989 I read this book by Marvin Minsky uh Society of Mind. I think it's called, which already promised all these amazing things like how a machine could make its own decisions and but it was all just graphics and it was really theoretical. So it took until me until like 2015 when deep learning kind of arrived on the scene and had all these amazing uh kind of results like, oh finally, you could detect the dog with 90% probability in an image. But I already knew this is just kind of you always have to kind of see the potential that's coming. And the moment I saw that this is going to be the next thing I jumped on for me, deep dream. Maybe you remember these psychedelic weird images. Uh That was the first time when you could reverse the process from just saying, classifying things to turning something like some numbers into an image. And yeah, so that was also the moment when it became clear that well, it could go beyond that. And uh yeah, I see really like, you know, grandfather's X, this thing where you start with an X and then you replace the handle, replace the Natalie Sharp in the end, like I tried to do that with myself saying, OK, which other parts and of course, that the thing like creativity itself and especially deciding whether something is interesting is good. Can that be replaced with a machine? And then when would you? So boto was born? Well, I mean, the idea for boto he conceive. So it came into the world in 2021. But I think the first time I wrote up the concept was 2018. And but as with any ideas that are maybe too early. I didn't find kind of, I have patterns. So, and sometimes I can go to them and say, oh, look like I have another idea. But in that case, they went for the memories of passers by first, which as we heard was made, eventually was sold at Sotheby's, which was nice. So, but yeah, that thing was in the drawer for another two years. Uh, it helps that then around, like with the pandemic, the Blockchain, the whole crypto world started exploding and suddenly people started seeing, oh yeah, there is actually something to it and then yeah, it took a year to build it. And yeah, in 2021 it bottle came into the world. Great. So now you're going to over the next few minutes give a a clear explanation of how Botto works for the layman and what you mean exactly by decentralized autonomous artist. So first, what does that mean? And how does botto work? Well, let me put this first. So really what I want to create with bottle or is like my answer to the question, can a machine be an artist? And especially now where everybody is becoming an artist by just entering some text. So that's the question like what is it? Is it just being able to produce pretty pictures that makes an artist? No, you have to create a compelling biography. And so that's how Botto is built up. So yes, underneath bottle is really a machine that produces images and tries to sell them. And by selling them, it makes money that keeps it alive. But there are some aspects to it. And I'd like to call the system more like the bottle Trinity. And I also like to call it like an inverse cyborg because this is how boto Trinity we're seeing on the screen. Yes. So explain simple terms how the Trinity works. The inverse cyborg idea is like unless like you do not have a human augmented by a machine, but you have a machine augmented by humans. And this works by these three parts. One is of course, the A I, well, whatever that is like the all the algorithms, the models that produce images and text. But then there is a human community in form of a so called DAO which is a Decentralized autonomous organization. It's kind of a, well, you could say it's company in the Blockchain, but it's a little bit different. But yeah, so community, it's a community that everybody can join and when you join it, you can make decisions and how do you, how do you become a, well, that's where the bottle token comes in or the economy comes in. So you have like just like, you know, Ethereum Bitcoin, there is a coin called bottle, which I call the lifeblood or the pulse and you can actually track the success of bottle in its chart. So you have to acquire bottle and then you stake it in a smart contract. It's kind of like you deposit it there and once it's deposited, you are part of the dow and you are allowed to vote on the books. They bought a small investment or a small, you can, well, you can join for free but you can uh the more you have stake in something, the more you can also profit from it. Because that is the kind of maybe the interesting part about botto. It is not so half of the proceeds. It may come to that later. With the process. We have a lovely graphic. So the process of how botto works, you have the dow which is influence it. So that's the community that teaches it. Is it art teacher boto go to art school. But the big question is like, why should the machine do any kind of art or what art should it make? So you could of course randomize it totally. But let's say realistically a fresh A I does not have any motivation to go in any direction. And that's where the humans come in that guide it. So bottle produces really constantly knew what we call fragments. It's like proto art. It's really, it prompts itself randomly by creating originally based on, on your taste. No. Oh, that's very important. No, no, no, no. Oh Actually, I have to take a step back here. I didn't want to influence the machine at all with my, so what it does is really like a shotgun approach to possibilities. So because that's what I believe is the best way to cover all possible directions. If it would only be my aesthetics, then, well, it would be a clone of myself and that was not the plan. It would be scary. Yeah. So bottle like uh roams through late in space and tries to get in any direction and builds up a huge pile of potential artworks right now. I think we are at five or 6 million that no human has ever seen. But once a week it selects 350 fragments. So thousands and thousands of fragments we boil once a week, it selects itself. So model using the taste model, it's like 350. Yeah. So it's like a curator in its own affairs, but that's when the voting comes in. So the humans uh get to see kind of like the proposals and can put votes on it. And in the end of the of every round, which takes a week, the one that got the most votes gets eventually minted on the Blockchain and set up for auction. That is when then there's the potential that it sells this and money comes back. But at the same time, those votes go back and trade in the taste model. So I'm the winner then learn that OK, this is popular. Actually, all the votes are all the vote. It's not only important what people like, it's also very important what they don't like. It's actually for me even more important because uh there are so many things directions you could go and we really create this kind of gravity of uh where it should show more and where it should show less. And unlike a human artist, it is kind of trying to please, right. A human artist should be kind of against the grain and really try to provoke some of the most successful try to please. Yeah, I mean, financially it's usually better to kind of go with what everybody likes. And if you, as you can see in bottles, art, it does a lot of things that are well agreeable. Well, let's look at some of the art. We have some slides, validating bottos success story over the last few years. Emerging press, we love. These are very early days. This is kind of from 2021 where there was not even these diffusion models like mid journey and all these available. So it was using a different technique. You can see a very clear kind of A I aesthetic, but I really thought out history aesthetic. Here. This is, yes, yes, yes. We have Edward Hopper. So here we can already see a little bit of a development when because what bottle works in is periods. So every period allows to add new models to it. And every period also has a theme which there's always 12 being proposed by bottle by the machine botto again, is deciding it and then the dow again, votes on what should we do next. And every theme comes with like a slogan and it becomes even like a prompt. And so it's like a several nested feedback loops here which and you can see some development that you can see how, let's say their quality, their style. So now it became more post photographic, I would say, and it's quite interesting to watch because not only the machine is learning here. What is for me is interesting, the human doll, which is kind of really people with art background, people without a lot of people who come from crypto, of course, and they all kind of pull in different directions and eventually, well, you get this essence and right now it seems they, they liked a lot, the photographic work. So you're seeing tastes change from both the human interpretation of votes, but also, but now 350 images such as this are produced a week for the Dow and the general public to vote, which is, which is interesting. So you don't have to be a member of the Dow. You don't have to have invested to actually vote on this. And um what we've got here, we hope it works is this week's or the last couple of days. Actually, the new round started the day before yesterday. I always mix up the time zones. But yeah, so every Tuesday uh is kind of where the next fragment is decided and immediately the new round star portfolio of work is there for the doubt to vote but also for the general public. And should I let's take a demonstration of where we are? So what you see is kind of like, I think you might have been given like information how to access it. So this is the the voting. It is really, you just scroll down. It's really like you're going through uh kind of like Instagram. And then right now I have 100 voting points which is kind of uh not a lot. So if you're a layman, if you just, but it's really to feel it out and then I can just say, OK, this is like for me, this is kind of boring. So it, it has typical A I cliches. So I give minus seven. This is also terrible. So really? So I don't know. So what you can see is that boto does not really have a single style and uh like usually kind of you would like to have an artist to have a very discernible visual style. What boto does have is a kind of, it has recurring themes and I find it quite fascinating to watch uh too, too typical. No, I hate it. So I'm, I'm more of the critical 10, do we get something loaded up? There we go. Oh No, no. Yeah. But I mean, there are these things like themes coming up, there's often like robots and machine hybrids. OK. Something typically abstract that might work nice on a kind of doctor's waiting room, I guess. Oops, I zoomed in and uh oh, actually I like this one personally need to get more of an idea of your title. And so yo you should note that, of course, botha creates all the descriptions for the for the works itself. Um And yeah, yeah. Yeah. So it analyzes them and uh like depending on what it sees, it, it never repeats any titles. So uh today the selection is, oh, actually, I like this. This is kind of post photographic glacial frame, kind of nice. So after the selection process, um many, many thousands of people voting, you have one winner. Yes. For the there's always a single winner. And yeah, that gets minted, gets put up for auction and if it sells and fortunately for the last 2.5 years, every single one sold typically, like, I mean, I don't know if you followed along the whole crypto kind of ups and downs. That was really kind of the winter and well, but still it went for about like 10,000 $15,000 every week. So in what happens then if it sells half of the proceeds are distributed among everybody who voted in that round, depending on how many voting points you spend. So, it's a little bit so the more stake you have in the game, the more you can win. But even if you don't purchase the artwork yourself, no, no, you don't have to get a share of the proceedings. Yeah. And it's quite interesting because, I mean, most of the people who are voting cannot afford bottos art, but you can shape kind of where boto is going to. So it is in your interest that itself. Well, and that's also the other question like, do you vote by your personal taste or do you vote because you think this will sell. So you have constantly this inner conflict dilemma, this dilemma like pity and commerce. Yes. Yeah. And, and that's what makes it interesting there. So, so an interesting economic model, 50% to the community who have voted and if 50% goes to Bo Treasury to keep the service running. Um Well, I mean, there is of course, and that's the other thing Boto does not try to kind of do kind of puppet play. So it's clear there are certain things that an A I at this point of time cannot like for example, take a call from Sotheby's saying like, oh, we want to auction it. There has to be still a human uh picking up the phone or answering the email. So there is a human team that gets paid and uh kind of, oh, I get a little bit payment even. But because Yeah, I have to take care for bringing in new models. But the nice thing is the more technology progresses, the less humans will have to do. So, we are now at the very future. That's what I'm intrigued about this. It's still very early days. So he, he's an infant, he's still in nursery, um the future. So how do you see it? It's becoming totally autonomous and able to vote for his own work. Yes. So one of the things is that both should be able to, to have more of a word itself. But what for me is the most interesting aspect is we have a potential to create a truly immortal artist here and not sure if that's a good thing, right? If you look at the art market, usually at least for the for the heirs, it's more beneficial. If the artist at some point leaves and stops making more work, it's always a good move young. Yes, that so bottle has the should never die. But yeah, what does it mean? So what I find interesting is that aspect of altruism in A I because as we can see the A I doesn't need all that money it makes. So why not share it with the world and do some good. And so for now, it might be just the people who are believing in boto and helping. But ultimately, the could also decide, OK, Boto should uh create a foundation like I don't know, if it becomes a billionaire, I hope it's uh able to also then think about other things than how it could change the world it would come. Yes. Move from being an artist to a politician. A I these are fascinating considerations. I hope that made sense. It did to me the second time around. I think it's a fascinating area, machines and humans and how taste is controlled or how taste is steered is fascinating and the future of this we will watch very carefully. So, thank you very much, Mario. Thank you. Thank you, Botto. Round of applause for Mario and botto. Thank you.

Vivek LUTHRA, Senior Managing Director, Growth Markets and ANZ Data and AI Lead, Accenture  Debanjan SAHA, CEO, DataRobot Moderator: Ellie AUSTIN, FORTUNE

Brainstorm AI Singapore 2024: Unlocking AI To Boost The Bottom Line

You both work with companies across a range of sectors. Can you each highlight one or two industries that you think are really leading the way in terms of turning A I integration into revenue? But let's start with you. Sure, thanks. Good evening everyone. My name is Z. I'm based in Melbourne, Australia and I run our accenture's regional business for the A I I think the key sectors where we seeing uh you know, business leaders that are what we call, uh you know, strategic bets include datas like banking, insurance, health, those three would be the top three sectors. We see a lot of action. OK. You're kind of similar. There are four industries where we see a lot of traction. One is financial services including banking and insurance, health care, retail and manufacturing. There is a lot of actually work in manufacturing where they're using both generative and predict act. And are there any industries that are encountering specific challenges? Do you think? Uh look, I think uh surprisingly, the industries which are more regulated are actually using it a lot. You know, contrary to what uh I think uh you would feel uh actually, if you see whether it's health, it's regulated, financial services regulated. I think it's mostly uh manufacturing industries which are, uh if you think about the spectrum of A I, you know, classical A I, which is in terms of, you know, let's say predictive prescriptive modeling has been around for a while. I think as we kind of start to join generative A I to it, we see for example, manufacturing companies taking a little bit different approach, taking time to be able to, you know, get it. But we are already seeing some initial leaders in that sector. And when you say taking a different approach, what are they doing? That's different? Sure, you know, the kind uh the context is a little bit different. Uh asset heavy industries have different types of business uh conditions in which they need to work. Uh If you think about to get maximum value out of A I, you got to get your data sets ready. Uh And for asset intensive industries, the data sets could be very specific to the type of instrumentation they use, which is not the same as let's say financial services, for example. So I think those would be the kind of thought process which some of the clients we work with are going through. So bon if you're looking at kind of your portfolio of of clients, can you outline some of the specific obstacles that you think exist to unlocking the bottom line as, as the session is called, well, I mean, you know, I actually see uh three gaps which I think uh our customers are kind of uh dealing with one they call the value gap right there are, you know, a lot of excitement about A I but translating that into business outcome is not easy for various different reasons. Sometimes there are not enough uh tools and platforms that people can use. Sometimes they don't have the data, sometimes mapping the business problem into an A I problem is not easy for people. The second thing that I see, I call it confidence gap. There are a lot of people who are kicking the tires, they are doing prototyping, but they are not confident enough to take those applications A I applications or models of production because they are not sure about the accuracy of it. They are not sure about the safety and other risks that associated with those models. And third of course, is uh the expertise gap, right? I mean, this is kind of a huge area of focus for a lot of people. Uh There are not too many people who actually know how to build, for example, foundational models, right? But luckily you don't need that many people to build foundational model, what you really need people who can use these models to actually solve business problems. Even that I think requires lot of expertise and training. Other thing that I think is also going to happen, people are worried about is that do all of this, that A I is going to disrupt the skill mix in an enterprise. And people are worried about how, what impact it's going to have. For example, you know, various different jobs that people do today, what is going to happen to those jobs. And there are new jobs which are going to be created by A I which requires also skilling up the workforce. So what would be your advice to someone, a young person who is thinking about what to study at university? They want to thrive in this new A I landscape. What should they be doing? What, you know, I have one of those at home. My son just graduated high school and he's a computer science major in um University of California and he's kind of a little bit worried what is going to happen to the coding jobs, right? Because, you know, there are a lot of coding jobs which I think this A I copilots are going to do, but I do think it's also going to create new types of jobs and where I think it's going to be probably most lucrative and productive in terms of skilling up is that when you can marry the A I skill with, for example, a specific domain specific skill, right? For example, if you are an expert in biology, how to apply A I to do new medicines or personal medicine would be a huge thing. I think even, you know, some of the other areas, for example, education, I do think personalized education will be a big thing. So how do you use A I to create those new industries, new use cases, new businesses. That's where I think going to be the most action and that's what people need to learn. So that's what people should be thinking about as they, as they map out their careers. I'm interested in the confidence gap that you the expertise gap, Rebecca. That's something that you, that you've witnessed as well. And if so, are there any demographic specifics to that? Is it, is the, is the confidence gap more prevalent in some parts of the world than others? Is it women rather than men? Are there any specifics around, around those factors? Go ahead. Do you have, I think, look, the we are going through a large pivotal moment. Uh I think not having the right talent in the right places is a challenge which every business is facing right now. We just can't find enough uh you know, uh talent to go about doing what we call reinvention, thinking end to end. How do you really not apply this incrementally? But actually invent examples, you know, we worked uh uh maybe a little bit sidetracked but we worked with the food and beverages company uh in uh in uh you know, globally and there by using an A I driven solution for marketing content generation, we could actually build one year equal worth of content in about eight days. And that is phenomenal in terms of productivity enhancement. But at the center of it is how do you actually transform your workforce to be able to really scale this? One needs to think beyond use cases to business capability. You know, how do you think about a product launch versus a photoshoot in which you can kind of apply something and so on and so forth? So I think in terms of the question you asked, I think talent is difficult to find. Uh companies have to think about what does it mean from a workforce transformation perspective at accenture, we specifically have worked a lot on what we call what the future of work is going to be starting from there to be able to kind of think through. So if that is the future of work, what is the type of workforce you need and then be able to define which type of workers with what skill sets are needed and have structured curriculums on? How do you actually transform the work that kind of becomes quite key. And I can show you today, I mean, we are trying to get talent in every market we operate in talent is limited. So it's a combination of getting the right set of basic capabilities and training them internally in your firm. Actually moving people, I mean, conventional data versus modern data is a big area. For example, because if you don't get your data, right, how do you kind of really unlock value from A I and so on and so forth? So hopefully that gives ideas. So in terms of confidence gap, I think it depends on the risk level of the use cases that you are addressing, right? I think, you know, the EU A I act actually has done a pretty good job in defining those risk level. There are use cases where you probably should never use A I, right? For example, social justice and things like that. And there are use cases which are high risk, there are use cases, there are moderate risk, there are use cases which are low risk and it depends on the regulation that people have. And you know, one of the things I tell people that um the CIO S when you talk to them, they are concerned about three things that can A I create top line growth. Can A I save cost? And how do I stay out of jail and stay out of jail is where the confidence come into play. And if you talk to the cio s of regulated industries, that's where they are most worried about because it can not only create business risk, it can create reputational risk. Sometimes it can create many other type of risk. They can be on the wrong side of the law, for example. So what's your advice? To the leaders in this room on bridging that confidence gap, getting to the other side of it. Yeah, I think, you know, the first thing to do is to define the risk profile of the use cases that you're working on. And depending on the risk profile, there are different steps that you take. For example, what kind of testing you need to do, what kind of disclosures you need to have, what kind of moderation you need to apply to the results produced by A I? Right? So if you think through that in a systematic way, I think you can take care of the risk and you know, figure out where you can take it to production and what actions you need to take. If things go wrong, I'm gonna open it up to the floor for questions in a minute. Um I've just got a couple more for you both. Um So Banton, you've been a data robot, a CEO for about 18 months now, I I believe is that about it? Um And when you started the company was, it was in a bit of a rough spot. Think there just been quite a few layoffs, but you've really executed a turnaround. There's a new platform 9.0 you're rolling out various integrations. Can you talk to us a bit about the role that A I has played in that turnaround internally? Well, I mean, you know, when I joined the robot, um gene TB, I kind of was not the thing, right? So Gener TB, I happened, I remember I was actually in a Morgan Stanley conference in San Francisco. And everybody asked me about Gener TB I and the way back, the Uber driver squeezed me for half an hour on Gener TB, I, I knew that Gene TB I has arrived. And so Gener TB, I has played really a big cut, you know, part in turning around the company because there is so much of opportunity that I see with Generative Bi and what we have also done, we have been in A I business for a long time. We have been a P A company for about, you know, 12 years. But some of the things that we developed for predict bi applies quite well to generative Bi, especially when it comes to governance and observable and things like that. And what I have seen is that is the combination of predictive and generative bi the intersection of that you find the most value added use cases for businesses, right? And I think of this in terms of the lib brain and the right brain, the lib brain is your analytic part, the predictive A I and the right brain is your communicative part. And when you put the two together, that's when you get the iphone moment of A I which is what has happened, right? Go ahead. The, that's a question for you. So accenture has committed I think $3 billion to A I um how are you going to ensure that that investment changes the way that you interact with your client in a positive way, maybe accelerate the process. But you also maintain that high touch human element that accenture is so well known for. What's the balance that? Sure. And thanks for going there, I think it's a very committed investment from accenture. Our CEO could have announced that in 2023. Uh I think the key parts to that was how do we kind of really work with the industry and shape the industry's future part of that was how do we build industry specific solutions, functional specific solutions, uh build a network of what we call uh you know, uh Ja I studios across the globe, be able to really help our clients uh see what's the latest uh across the globe when they come to such a center. Uh The other part was uh you know, uh training our own workforce trying to kind of do that. So largely how do we keep humans at the at the center of this? If you look at most of the research accenture has published quite a bit, about half of the benefits come from automation to the bung point as you go from classical A I, which was all about data to generative A I which is so much about language when you put it together, what you can do in terms of automation, like, you know, autonomous processes that scale is amazing. Uh However, you know, as you kind of go towards, you know, putting this together rest half of the benefits are what we call augmentation based. So you got to start thinking not incrementally around processes, but you need to start thinking reinvention. How do you actually start to think in a new way, how you would do customer care and it's very, very different than looking at how you're doing it today versus how do you want to do in the future for doing augmentation? You got to start with putting humans to the center thinking about in this new world. What is the role which machine plays to kind of really support humans and still get to benefits and the benefits are not just automation, productivity benefits. The real benefits are around outcomes having different types of work, which is more interesting getting to what we call a 10 X impact in clients, business and so on and so forth. So I would think that's how we are looking at. How do we put humans at the center? Quick question before I open it up. The bon you mentioned before the iphone moment for a I, I'd love to know from both of you. What would be your personal iphone moment for a I maybe it's happened but maybe it hasn't. What would that look like for you? That light bulb, massive light bulb moment? That changes everything. Well, I mean, you know, it is when A I becomes pervasive in everything that we do, right? And I don't know if I'll call it iphone moment, but this is an example use case which kind of fascinated me for a while. Uh So we have customers who use classical A I to do various different kinds of forecasting. We have a grocery store customer, they own a lot of grocery stores in the US. And they use our models to predict, for example, how many mangoes they need to order in a particular store in California, right? And they have been doing that for every product in every store. Some various different days have different models and there are thousands of those more and what they have done with Gene Bi and which is quite fascinating. Now, they figured out how many excess mangoes they are going to have in a particular store. And they are creating personalized coupons that they are sending out to people that they know they are also creating, for example, very, very localized radio ads that they are putting in various different radio stations. For example, if they know that Indians like, you know, ripe mangoes and Mexicans like green mangoes, they will put a Hindi Language ad in a local news station and they are doing all of this thing automatically. So, you know, pretty much everything that we do probably can be made more intelligent with the use of A I and it's when it becomes pervasive and invisible, that's when you know that iphone moments will happen. So, no leftover, man, guys, that's very interesting. It's interesting. Look, I would kind of want to say that maybe last year it was mostly about experimentation and we work with businesses, but this year we are seeing businesses really put their heads together to be able to scale, to be able to scale. You need to think about, you know, the competencies across having a large language model is a small part of that which you use for doing whatever. There are normally seven things we see leaders who do this well, kind of bring a lot of things together to be able to scale this up. So think about being on cloud, that could be one thing. Second is getting your data ready, which we spoke about third, think about the enterprise grade A I solution. That's not just about the model on the model, you have adaptation on top of adaptation, you have consumption. So you've got to put that thing together, which becomes important. You also want to integrate that with your, you know, digital platforms, like your enterprise, the P platforms, the marketing platforms, the sales platform. So there are a lot of these things which come together to be truly be able to unlock the value and be able to scale up. And I think that to me is where we're seeing a big difference from last year to this year. Any questions for the V and a man? And if I raise your hand, if not, I will continue with one last question, I, I wanna finish on the subject of timeline. So you've just out laid out multiple steps. So that um if a, if a client comes to you and says, hey, we're about to invest a ton of money now in A I, this is our priority going forward. What's a realistic timeline for them to see meaningful returns? Sure. And, and let me take uh examples, right? Uh I think in the press uh uh accenture and National Australia Bank recently announced a journey we've been on with them for about 1.5 years now. Uh through this process, they focused on how do we invent uh uh using uh A I generative specifically in customer service and operational efficiencies. Think about 20 odd use cases, about eight of them being enterprise grade level. So that's about 1.5 years. What I want to say is I think the key thing is it is a multiyear journey, right? If you're thinking about reinvention, my encouragement to leaders I work with is to think beyond use cases to be able to go at scale, you got to move from use cases to business capability and that takes time. Take a lot of thinking, I would highly encourage all of us not to kind of really get bogged down by incrementally changing the process investing in your own workforce is going to be very important. And then thinking through all these seven conference we spoke about, but really it is a multi year journey. You know, you got to put, be thinking, how do you not going to do incremental changes to what you're doing today to be able to and one final tip for the people in this room, if they, you know, they in companies that they want to really unlock as we at the bottom line going forward. What's the one thing they should go into their offices tomorrow and start thinking about? Well, I think I do think use cases and returning value to business is important. I mean, you know, for J we are in the honeymoon period and that's not going to last very long, right? So it is important to pick the use cases which show near term value, but also have a long term plan so that you can show the value at scale, which I think is going to happen. But showing near term value and showing some return on investment, showing some successes, think it is very, very important. One final bit of advice pretty similar, I think lead with value. That's what we do with all clients. You've got to be really knowing what you're solving for from a business perspective. This is not a tech for tech. After decades of technology life cycle, you have something where the technology is adapting to how humans work. I think it's an exciting era if approached correctly. There is so much to be done in terms of unlocking value. Thank you so much for joining us. Thank you. Thank you very much.

Otto

Otto in Fortune Rankings

Videos