International Business Machines Corporation (IBM) Earnings Call Transcript & Summary

Hello, and thank you for joining our webinar today to talk about how to improve external data sourcing and scale to meet AI demands with weather data. My name is Antonietta Rubinetti, and I'm the portfolio marketing manager here at The Weather Company. Today, I'm happy to welcome Bobby Cameron who is Vice President and Principal Analyst Serving CIO Professionals at Forrester. Bobby will be sharing some market research and latest trends around the use of external data sets with AI. We also have here today Shikha Garg, who is the Offering Manager for Cloud Pak for Data at IBM. And Shikha will be presenting some interesting information on how you can use the IBM Cloud Pak for Data as the unified analytics platform to help you simplify and automate your journey to AI. And then last but not least, we have John Bosse, our Offering Manager of The Weather Company, who will be talking about how you can get access to weather data from The Weather Company and also some essential use cases on how you can use weather to help bring some weather insights into your operations and inform your business decision. But before we begin, just a few housekeeping notes, this webinar is being recorded and will be available on demand. Also to give you a quick tour of the webcast console, you should be able to see on your screen the media players and window for slide and the Q&A widget. So if you have any questions during the presentation, please feel free to type them into the Q&A window, and those will be answered at the end of the presentation. So without further ado, Bobby, I'll pass it over to you.

Great. Thank you, and welcome, everyone. Appreciate you taking the time to speak with us today, and look forward to hearing what your questions are. We thought it'd be appropriate to start the conversation with a very specific reference to the pandemic that we are all working our way through. The reason being is we're going to talk about trends, we're going to talk about investments and the like. And not everyone is in the same situation in the middle of the pandemic recession as we were a few months ago or as we might be in a few months after we come up with some solution, a treatment or a vaccine. So this slide attempts to communicate the 3 different modes that we see the pandemic having on companies. The survival mode, that's 23% of the companies; adaptive mode, 58%; and growth mode, 16%. Then you see along the bottom there, the green strip along the bottom, 10% or more revenue loss is survival mode, holding flat or maybe reducing -- seeing reduced revenue of up to 10%, and then growth is growth. And the key message here is that you and your company, if you're large enough and complex enough, may have all 3 modes [indiscernible], and these modes are going to shift over time. An example of a survival mode would be Scandinavian company operating 10 airports. They're down to 5% of their total revenue. They have 95% less revenue. All of their passenger revenues disappeared, and it's all cargo. So very low-margin and very difficult to make money. So the whole IT strategy there is essentially put on hold. The growth mode, you get companies, let's say, some of the direct materials associated with treating COVID, so masks and hand sanitizers, those sorts of things are in very strong growth mode, accelerated by the pandemic. So you can understand that a little bit how it works. So as we work through this, I'm encouraging you to listen from the strategic point of view, our data we took a sample in December, January, turn of the year, and we took a reconnect with about 1/10 of that sample and in May. So companies were 3 months into the pandemic, and we found very little shift in strategic intention just that budgets weren't always there based on the mode. So that's the reason for starting out with this slide. So keep that in mind. Now today, we're here to talk about AI and talk about external data. And I thought it would be important to start with a little bit of a view as to what's going on in companies as they try to address the whole question of data and the rising importance of data. And this particular slide shows the context of the chief digital -- sorry, chief data officer. We found in our last survey set, which is the end of '18, so a little over 1.5 years ago, that 50% of global data and analytic decision-makers identified that their firms had a chief digital officer. And those report about 1/4 of them into the CIO, about another 1/4 into lines of business. And others report actually up into operating execs like a president, CEO. And these -- what's important about this particular role, it's part of a shift to the importance of business more than technology, and you'll hear me make that point a couple of times, that as we integrate the planning and how we're going about using data, the challenges we have are much more about how to use data for differentiation. So what can we do with it? It's about better business insights, driving, having C-level champion, so we can drive an integrated point of view. And so we see the rise of the chief digital -- data officer -- I keep wanting to say digital -- chief data officer, and the focus on the business challenges as a key piece of what's going on. And by the way, the bulk of the firms who don't have a CDO have -- and even some that do, have a center of excellence with which they're trying to manage the same problem. But the CDO, chief data officer, creates an opportunity to actually drive focus. And that's the reason that this is an important concept. Now let's drill into AI itself and see how that's working out. And I find it very difficult to talk about AI without spending a least a little bit of time on the breadth of the definition. AI has so many different technologies rolled into one. And all too frequently, we get casual about how we talk about it. So here are the 4 categories that Forrester uses in our research to try to differentiate. So sensing. That's the whole process of speech recognition, speech generation, text, image interpretation. So it'd be facial recognition, for instance, video interpretation, lots of different ways to use the AI to interpret bitstreams of information. Now thinking this is a more common form of AI, it's machine learning, deep learning, where the effort is to use the technologies to actually increase the understanding of the data. And this is a critical piece when we start looking at external data. But also, it's critical piece of the integrated view across the company. And I'll herald the message around the external data here is that we see companies wanting an integrated view of the customer, let's say, or trading partners or markets, and we could get from our own data, pretty good view of the inside-out perspective. But it's very difficult for us to understand what's going on in the ecosystems around us without that external data. And that external information then provides us the ability to interpret it and apply the thing. Now acting much more, as the word says, much more tied to specific outcomes and generating results. Now here you might have search, speech, speech creation. It might have that in chat bots, for instance, those sorts of things; smarter decisions; driving specific information, for instance, big data, software robots and an outbound finished goods supply chain, optimizing distribution. So that's applying think to drive action and moving inventory shipping plans around in order to optimize the data. Foundational. Now these are the engines underneath. This is to power the data, big data sets, integrated views, aggregated points of views and the applications to manage the data and exchange it, and the infrastructure to operate the AI. And a nontrivial piece, which is the consulting services to make all this work. Now I spent a little bit of time talking about these characteristics to set up a little bit more of a drill down now into what are the challenges when we step into AI, and in particular, a lot of those challenges show up around automation. And here, again, I said I would herald that technology is usually not the issue. You can see in this survey set that the dominant challenges come from people, leadership and organization. They outweigh the technology leadership. And what that suggests then is that as we are planning -- as particularly a lot of us on this call are technologists, as we're planning for what we're doing to improve artificial intelligence, whether it be around actions and automation or maybe driving some chatbots, some other things like that. We have to pay attention to training and skills in our own technology organizations as well as through the nontechnical staff. For instance, you might be working with software robots, et cetera. We need to worry about budgets. One of the reasons the chief digital officer shows up, by the way, is -- specifically around AI, is because the skill sets that we're talking about and the kinds of challenges that we're looking at here don't just occur in isolated organizations or for 1 use. The more integrated the use of the data is, the broader the set of organizations that we use -- that use the data and are trying to take smarter actions. And putting that in the context of the external data, the problem even accelerates, and that more and more parts of the enterprise have an interest in that broader view of the data. So now let's drill down a little bit here and shift into looking at the external data and what's going on with that data. And here, I've tried to summarize a lot more words than chart, but there's a lot of conceptual stuff that's important here. We'll look at that, the external data, from 2 perspectives. From here is some data -- some percentages around utilization. Pretty impressive utilization of external data. And the next slide will characterize some of the uses that are underway. And particularly looking at that first major bullet, the -- most firms are expanding their ability to source external data. 56% at the time of this particular survey, with 20% more that they plan to do external sourcing in the next 12 months. So that says that they're -- that's 68%, 2/3 of companies, are in some stage of significant use of external data. So it's a nontrivial piece of that integrated model and applying AI in order to pursue the data. And again, I'll show you some of those utilizations in a minute. Acquisition itself is a very interesting piece. It's starting to -- the acquisition's starting to span across the company. And that acquisition starts to show up in lines of business, shifting control out from IT. You can see the 2018 over 2017. So the '18 data again being about 1.5 years old, down points from the year before in terms of numbers of companies where IT is managing it. A lot of that's because the utilization of the data spreads out. And again, that's a driver for the chief data officer role. New roles are emerging. One of my favorite is that the data hunter, someone whose job it is to go out and look for external data that helps improve operations of the company, improve decisions, et cetera. So a nontrivial piece of change. And then finally, insight service providers. Are there third parties out there who are hosting the data? And we'll hear from IBM around the weather data in a minute, it's a great example of that, where a lot of the analytics is done on the data, and the results are made available out to these people who are buying the external use. That model works very well in a variety of modes and is much more effective and efficient than trying to ship all the data around to move it all into one place so you can have a common orientation. So we've been looking at what's going on in the organization structurally, with more and more pan enterprise ownership of data and how it gets used and are focused on integration. Then bring that into the AI world and how the use of AI is increasingly an organizational and people challenge so that we got a -- as we gather more data, include the external data, we got to be smarter about it. Now we can see some of the characterizations of the growth. And we'll finish up, before I pass it along to Shikha, the -- looking at the kinds of external data that companies are pulling in. There's person data: information about customers or trading partners or markets; a place data: that's looking at various kinds of environmental conditions, maybe traffic, certainly, weather, maybe location of specific businesses. Starbucks is one of my favorite to look for, or used to be when I traveled. And then things. The Internet of Things posing a wonderful opportunity for -- in lots of ways, both in terms of individuals. In the middle of the COVID, you can find -- Germany, for instance, are actually bypassing some of the GDPR constraints around personal data as people are using their smartphones, talking to each other and tracking exposure in case there's a need to trace connections that someone may have had, who is diagnosed with COVID-19. It's a very interesting use there. But a lot of other analytics going on using sensors and the like and analytics that go out at the edge of the network. So you've heard the logic of edge. Clearly, I think the Internet of Things is accelerating as a huge piece, but the other parts as well. So when you think about your data situation and you're looking at the technologies and the data that you're working with, think about this complete picture. Oh, and I did not flip the slide, so you're saying, what is he talking about? Here, just -- I'll leave it for just a moment, Shikha. Here are the 3 types of the data. And this whole perspective then is about how you use the data and how you engage that external data. So let me pass it along to you, Shikha, and I appreciate it, and we'll -- I'll be around for the Q&A at the tail end.

Thanks a lot, Bobby. So hi, everyone. This is Shikha Garg. I'm Offering Manager in Cloud Pak for Data. So as Bobby explained, right, that how enterprises are trying to create differentiation using data and artificial intelligence, we see there is a trend that where clients are already seeing business results by adopting a common platform for their analytics and AI solutions, from improving customer experience to help employees be more productive, or to resolve security incidents faster, right? So to achieve all these results, a common platform is what is helping the enterprises. And that's why IBM has Cloud Pak, which helps provide different kind of unified platform to achieve the objectives. All the IBM Cloud Paks are enterprise-ready, containerized software solutions that give clients an open, faster and more secure way to move core business applications to any cloud. So each Cloud Pak basically includes containerized IBM Middleware and a common software service layer for the development and the management, which is designed to reduce the development time up to 84%. And if we dig deeper into the Cloud Pak for Data. Basically, that's the enterprise analytics platform, which IBM provides, where our approach starts with a very simple idea, right? So you build once, and then you can run anywhere. Our platform can be co-located where your infrastructure investments have been made or will be made. And so we support deploying the solutions, build on Cloud Pak for Data on top of every major cloud platform, including Azure, AWS, IBM Cloud, and you can deploy it in on-prem based on your hybrid case scenario as well. And the way we support all this infrastructure is by layering [indiscernible] and then reduce a common layer of integrated services that allow you to collect information from any repository, any database, data lake. And our intention, basically, is for you to leave the data wherever it resides. Our collect layer basically introduces the capabilities such as data virtualization that allows you to fold multiple database schemas into one, and this allows you to eliminate the need of moving the data back and forth and do various kind of ETLs. So once the data is collected, that's when we have the organized layer. Again, keep in mind that all of this is available in the same platform. So your data engineer, your data stewards, your data scientists, all these user personas are working in the same platform rather than in disjoint applications. So once your enterprise data has been collected, we have industry-leading data cataloging capabilities where you can create a single place for your data engineers and data scientists to shop for the data. And once the data has been organized, then we bring the function of data analysis to this enterprise catalog, and we provide a sliding scale for our data science capabilities so that user can quickly grab the data and do whatever analysis they are trying to do to solve their business problems. And this is a sneak peek on different capabilities, which are available as part of our Cloud Pak for Data platform, where we provide the capabilities from data virtualization to instantiating various databases, using various SQL and no-SQL databases. And for real time, we have our Db2 events tool and streaming analytics available in our collect layer. Similarly, we have various data transformation and cataloging capabilities in organize. And eventually, in analyze, we have our whole Watson portfolio to build machine learning models, to deploy it and to eventually create various insights based on your data analysis. So what we have done is that on top of Cloud Pak for Data, given that Cloud Pak for Data is already an enterprise-ready analytics platform, we have -- and this is the survey which we did with Forrester, where there -- we did that how much basically return on investment you'll get by implementing a unified platform such as Cloud Pak for Data to reduce your development effort and to reduce your infrastructure management effort. So what we've done now that on top of Cloud Pak for Data, we have brought the access of external data. And as Bobby was mentioning that the importance of external data is paramount more so into a scenario where the running analytics on just enterprise data is not enough because you need to understand the holistic view of your consumer, how their trends are changing, what kind of demographics they do live in. So how do you integrate this external data back into your enterprise data to do more broader and 360-degree view kind of analysis to create the differentiation on your offering, right? So that's why we have worked with various industry leaders of data brokers to bring the access of these external data sets available on Cloud Pak for Data. And these data sets can either be available as an offline data set, where if you are in an on-prem environment or you are -- you do not have interim access for your analytics platform, then you can download the data and that access, make it available for your data scientists, or you can use APIs for real time access of the data as well. Again, this is a sneak peek of what kind of data providers we have available today, and we are continuously working on increasing various -- more data providers available on our platform to create a marketplace which our clients can use seamlessly while doing various kinds of analytics on Cloud Pak for Data. So one of them definitely is The Weather Company, and John will talk a bit more about once I'm done with this slide. Apart from weather data, we have labeled unstructured data for images, videos and audio available as part of the platform. And then we have partnered with various companies, such as Equifax, People Data Labs and BCC Group to provide various other kind of premium data sets, right? So for example, Equifax provides you the demographic information, the credit information and the financial capacity of their consumers. People data provides the personal information of a consumer. And the BCC provides 1 single API to get real-time financial data. So these are different kind of data, which based on different use cases, which you're trying to solve in your business unit, you can help upend your enterprise data with these data sets and create more holistic view of the insights and create the differentiation for your offering. So John will now talk about more details about The Weather Company. John, back to you.

Thank you very much, Shikha. Yes. So weather, as Shikha mentioned, is included in the base of Cloud Pak, and hopefully, that inclusion is obvious, right? Because weather certainly affects every person on the planet. And as such, it affects every business on the planet. There aren't -- nobody is immune from the effects of Mother Nature, and certainly, understanding how weather impacts the business can be quite important. And becoming more important now because climate change is really increasing both the frequency and the types of weather extremes that we're seeing. And you don't have to go too far back in memory. Just earlier this year, the devastating wildfires in Australia. We are seeing crazy changes in the climate, and those changes are impacting businesses in very different ways. So from temperature and precipitation extremes, we're seeing droughts and wildfires that might last longer. Temperature extremes becoming more common. And then events like floods and intense precipitation in areas can cause problems. And that's on top of common things like hurricanes and tropical storms like we saw in the East Coast of the U.S. last week, or blizzards in the wintertime. It's more and more common, and businesses have to plan for that, have to understand it better, and have to understand how it's impacting them. Now the Institute for Business Value did a study a few years back of 1,000 C-level executives around weather. And the numbers here are pretty amazing. 100% believe that weather impacts at least 1 cost and 1 revenue metric in their organization. 93% believe that improved weather insights can positively impact their annual revenue growth. Now that's amazing. And those that are successfully embracing it are seeing it. They can prove those points, right? Norfolk Southern, who we work with, has reduced switching costs and carbon emissions by 70% by employing weather information during their winter time operations. One World Observatory, just a retail-type example, is optimizing their resources, their staffing to have the right staff on hand at the right times to meet their crowd demands that they're going to have. Unfortunately, there's a lot of companies that haven't embraced weather yet, right? They haven't figured out how to implement it to make systematic decisions with it. And what's holding those companies back? We found that 69% aren't certain how weather data can create value, 60% have difficulty integrating it into their processes, and 57% aren't even certain how weather is going to impact their decision-making. So that's quite large numbers for kind of inactivity, right? Not taking that step forward and embracing the obvious. So what we want to talk about is utilizing that weather data that's included in the base for Cloud Pak for Data. How can you achieve insights from that? And it's really quite simple. If you step back and think about your operations, you could probably pretty easily identify areas within your business that are most impacted by weather, whether it be SKUs in your business, whether it be processes in your business, equipment. Once you understand what is impacted, then you can work to quantify the impact; not just the impact that weather has on that, but the financial impact of what it can do to that business value. And then once you understand it, you can start to predict, you can utilize weather forecast and weather understanding to gain insights into what your risk may be in the coming weeks or even months using some weather insights. So what is included with Cloud Pak for weather data? What we've done is we've packaged up access to our weather APIs. So it's a very streamlined, highly scalable, reliable access to data in our Weather Company data limited edition, which is available with Cloud Paks. It gives you the API key for 90 days where you can utilize historical weather data, current weather conditions, forecast conditions and location lookup services to really understand kind of what's going on in any particular location. In addition to that, if you find value in that data, you could supplement what you're getting from Cloud Pak for Data with premium services. So you can get, whether it be industry solutions or specific things that we may have built, The Weather Company for industries, such as energy and utilities or ground transportation or agriculture. We have weather alerts available, so if you want to trigger certain activities based on weather conditions, as well as geospatial data, dashboards to help visualize it and even enhanced APIs that could augment some of the analysis that you've done. So let's take just a quick look. I'm not going to go into these in great detail. But analyzing when the data is done today across many different customers that we have. So we have retailers doing at the store performance, understanding their point-of-sale data, product categories and how weather data impacts that so they can maximize sales opportunities or reduce extra inventory. It's all available, and it's being used, like I said, by a number of different retailers today. The supply chain efficiency, really understanding how weather is going to impact the flow of goods between operation centers, distribution centers, manufacturers, understanding that, avoiding those problems, making sure you have on time deliveries, inventory where it needs to be placed. It's also used in for preventive maintenance. So really understanding your assets, that asset location, but how weather impacts that asset. Once you have that understanding, you could understand -- you can utilize weather data, forecast data and such to predict when maintenance might be required. It could be acquired ahead of schedule, if it's been in more adverse conditions. It could go a little bit longer. If it hasn't been in adverse conditions, you can have a better picture of for your digital twin of that as asset. And another example is in insurance and claims processing, understanding the call volumes you might have in your call center with various events, staffing at the appropriate times for those events, and also placing your adjusters in areas that they need to be to deal with various outbreaks. So that was just a touch on some of the use cases. So we talked about weather data, we've talked about how it works for Cloud Paks, we've talked a little bit about use cases. Hopefully, that doesn't seem too daunting. But if it does, Shikha and her team have worked hard to provide a couple of accelerators along with the Cloud Pak. So utilizing these accelerators, the sample notebooks, will help you more quickly import your data and get some tangible results, do some analysis quick and understand, maybe identify some quick areas where you could implement processes to improve your operations, utilizing the weather data. So the Cloud Pak comes to those accelerators to really give you a jump-start that you might require. So with that, I'm going to finish and turn it back to Anton, and thank everybody for your time today as well.

Thank you, John. Well, some key information here for you if you want to get started today. First of all, we have here a few links. We're going to share this presentation after this webinar, where you can download the Forrester's report on total economic impact of IBM Cloud Pak for Data. You can also check out these new industry accelerators that John just mentioned. So these are on the developers hub, you can see some examples of sales predictions with weather, retail predictive analysis, also some manufacturing examples, we are adding some insurance risk analysis. So there's going to be more of these industry accelerators posted on these websites for that to help you kind of find out what is the right use case for your company. But again, just make sure that you have access so that this, 90-day trial, this link actually brings you to the page where you can request your API keys. Once you have your API key, you get immediate access to these weather data package for Cloud Pak, and you can start kind of exploring and doing some of these analyses on your own, and you can always reach out to us to upgrade any growth packages. More information, here's a link to our web page. So I'll leave you here as well with the contact information for our speakers. Feel free to reach out if you have any additional questions. If not, through our web page, you also have an opportunity fill in the contact form, and we can circle back and reach back to you.

So with that, let's move to Q&A. So we have a couple of questions here from the audience. But before we get started, I just want to remind everybody that you are on mute. If you would like to ask a question, you can do so via the Q&A window. And we will do our best to answer during this call. If not, we can always get back to you. So let me move here real quick and can go through the first question. I think the first one, actually, Bobby, is for you. So if artificial intelligence covers so many different technologies and approaches, [ what is the ] use case for investing in AI?

Well, that, I think, is one of the most -- one of the strongest reasons we see the chief data officer emerging is that the utilization of these technologies span the enterprise, and with the external data involved, reaches out into the ecosystem so that the conversation around funding the technologies to apply to the data as well as bringing in the talent to actually do the work starts to span multiple funding entities, if you will, multiple lines of business or multiple operational functions. And so what companies are doing is actually starting to address something like AI as platforms. And they might be broken down by the style. So we were looking at since, for instance, and act and think as being styles. Or they might be breaking -- broken down into platforms, thinking about how the data are used. So the person, the place, the thing. And these platforms then, because they're a larger scale. They're not individual technologies, not a specific database, not even necessarily a specific technology, can be characterized as how they're supporting the business. A lot of the time, they'll be used for differentiation, and that was in Shikha's comments about using Cloud Pak to differentiate among the different companies who are competing. And by the way you use the data, how you intend to use it, you can build out a case for -- not for a technology, but for a family of technologies, that then allow you to apply the AI and the data in order to achieve some sort of differentiation impact. And an example of what that might be would be understanding market changes quickly, being able to -- well, John gave you some great examples of using weather data. And if you've got some specific business activity, or it might be shipping and distribution, let's say, and you're watching how weather events are making it difficult to deliver, you can take some specific actions. Those might be ways you differentiate so you've got a more intimate relationship with your customers and crisper definition. So the bottom line recommendation, instead of trying to tackle AI as individual technologies and the data as individual data sources, look at the aggregation of those and think about the benefit it brings, especially in terms of differentiation.

Okay. Great. So I have another one, I guess, John, for you. Where should we -- sorry, which versions of Cloud Pak are entitled to use weather data?

Oh, thank you. That is a good question. Let me clarify that for everybody. The packages Weather Company Data Limited Edition, and you can find that to be used with both Cloud Pak for Data and Cloud Pak for Integration. The weather data can also be used both inside and outside of the Cloud Pak environments, but it is required that the client at least be utilizing either Cloud Pak for Data or Cloud Pak for Integration.

Okay. And one on the -- where should we store our data? External or internal? And why? I think this is something Bobby and maybe Shikha wants to add her perspective to?

Sure. Let me start, Shikha, and then pass it to you. The thought there, just real quickly, is that you don't want to spend your time moving the data around. So I mentioned, for instance, in the -- using things and their location and characteristics that are going on. So maybe sensors at an edge -- in an edge environment watching -- let's say, watching traffic. And that kind of data, you want to flow all the way back to some compute environment. So you actually move the translation of the data, the application of artificial intelligence, too close to the data. And that's actually one of the benefits of working with a cloud provider who gives you lots of locations. So you end up being in proximity to a compute environment that maybe is more minimal. So you're not moving data a long way away and you may actually put compute engines at the point of contact. Shikha...

Yes. Just to add just one more point, as Bobby said that, yes, one is that you move your data storage next to your compute, right? And don't move everything because you just need to filter down and create a definition that -- what would you need eventually for your usage, right, because you don't want to store like every second or every millisecond of data coming from your IoT devices. But on top of that, you should use definitely the data virtualization capabilities from whatever cloud provider or software provider you are using because the virtualization is what basically defeats the purpose of moving the data, right? So then using the virtualization capabilities, your data can remain wherever it is. And using the advanced capabilities of the virtualization softwares these days, you should be able to reduce the whole purpose of moving the data, creating a data lake, and just keep the data wherever it is and be able to do analysis back and forth from wherever the data resides.

Okay. Great answers. Thank you. And there's one more here. What should we do if we're interested in using weather data beyond the trial? Maybe you, John can answer that one?

Yes. I can certainly take that one. So yes, as I mentioned, the package includes a 90-day trial, and that isn't limited to just test environments. You can use it and test in production environments. And I believe the API access is limited to 50,000 API calls per day, which is roughly 100 calls per minute. So pretty hefty use is available on that. But beyond that trial, beyond that 90 days, if users want to purchase that, The Weather Data packages are available through The Weather Company. Access to it varies by kind of the API type. We have a number of different packages available. So if you need historical data for access, or if you need forecast data for prescriptive and predicting, it's available. So you can contact your IBM contact for The Weather Company, or any IBM contact, and they'll get you to the right folks to figure out what data package best meets your needs and get you a subscription to that.

Okay. Thank you, John. And again, just a reminder, if you have questions, please make sure you type those on the Q&A window. And we have one more here. And it's, where should the CDO report to? Into IT? Or business unit executive?

That's -- I'm going to take that one. The -- I think the concept of the CDO is an important part of the company growing up and maturing its use of data so that we find them more common in companies that are beginning their data journey or in the sort of the middle stages. And the reason they're there is to pull resources from across the enterprise and optimize those resources against the overall demand and interest in AI and in data. And because of the complexities we've been talking about, location of the compute, access to specific sets of data and how all that gets translated, let's say, into a business action, all of that gets -- is pretty complicated when you get started. Now -- so therefore, the CDO, chief digital -- data officer reports more senior in these earlier stage. As these companies get stronger and stronger, they tend to have that chief digital -- chief data officer, spread out more. And I'm thinking about digital here, I was thinking about Allied Irish Bank, where the gentleman who was in charge of data and reported up at that point to the CIO, a very senior role, during their transformation to digital he became the chief digital officer and data was part of his focus. And then after they completed about 9 or 10 years of digitization, then the data and digital, as a specialized activity, was rolled back under the CIO, but there was no longer a CDO, data or digital. And that example points out that the more mature a company gets, the more of these roles around data and how you use it become part of everyday business. They're not unusual. They're much more integrated in. So quick answer to the question is that the CDO in the start-up phases, earlier phases of artificial intelligence and data, and the increasing use of the external complex data, it's a business problem to be solved that spans political boundaries and therefore requires the chief data officer to report more senior. The more mature the company gets, the more the digital and data activities get rolled out into the company and, therefore, become part of more of the common day-to-day business. So the less likely to even find a chief data officer.

Okay. Excellent. Thank you for that explanation. So well, it seems like we have answered all the questions for today. I would like to thank you all for joining us. And thank you to our speakers, Bobby, Shikha, John, for the great presentation. Please expect to access this webinar recording within the next 48 hours through the same link. In the meantime, feel free to reach out to us if you have any additional questions. And with that, I would like to thank everybody for your time. I hope you have a lovely rest of your day.

Okay. Thank you.

Thanks, Anton.

Thank you. Appreciate it.

Bye.