In this episode, Jo and Maribel have a conversation with David Dobson, Global Industry Director Retail, Hospitality & Consumer Goods for Intel, about the Value of AI at the Edge
In this episode, Jo and Maribel have a conversation with David Dobson, Global Industry Director Retail, Hospitality & Consumer Goods for Intel, about the Value of AI at the Edge
Hello and welcome back to the podcast. I'm Maribel Lopez and I'm joined here with my ever lovely host, Jo Peterson. Hey, Jo.
Jo Peterson:Hey, Miss Maribel.
Maribel Lopez:All right, and we are here today with David Dobson. He leads the global retail hospitality and consumer goods industry strategy at Intel. And he's coming to us from the UK. Hey, David,
David Dobson:Hi Maribel. Hi, Jo, nice to meet you. And thanks for inviting me to the podcast.
Maribel Lopez:Thank you. So David has been with Intel for more than seven years prior to his time at Intel, David spent over eight years at Microsoft working in retail, hospitality, wholesale distribution industries around the world. So deep bench of expertise in a lot of the industries that are very edge heavy. And today we're going to be talking about the value of AI at the edge. So, David, from the data we're seeing, with the use of AI, it's dramatically increasing in the enterprise, particularly in certain verticals, the ones that you're familiar with, in particular, retail banking, hospitality, we've seen growth rate numbers that go anywhere from 10 to 30, maybe even higher in terms of adoption of AI in those fields over the next five years. So as you're chatting with customers, what are you hearing and seeing in the sectors that you think we should be aware of?
David Dobson:Yeah, I know, we'll get into a bit more details later. But I think I'll start by saying this is the number one conversation we're having with our customers, I think the past the stage of what is AI, and they understand the technology, but really most of the conversations we're having today is around, or how can we best leverage how can we best use this technology to drive our business. And in that area, I think what's making it difficult for a lot of these businesses is they require multidisciplinary teams, right. So it's not just about some technologists that can sit down and magically just come up with a great solution for a particular problem. You need people who understand the problem domain that you're looking at. So in retailers, that's maybe people who understand deeply what your supply chain looks like, or deeply how your store operations look like. Or in banking space, it could be people who really understand fraud and risk type calculations. So you need those deep domain experts, you obviously need the people who understand the artificial intelligence technology. And then you need the people from the operations side who can operationalize some of the work that you do. So it's kind of a multidisciplinary problem that they're trying to fix. And really how to bring those teams together. And how to use this technology to apply it to some key business areas is really the big conversation that that we have with many of these these businesses today.
Maribel Lopez:To David, let's double click on that a bit and talk about the use of AI in the banking industry. As you noted, most financial services institutions have been using at least certain aspects, certain aspects of AI for things such as fraud detection, and really many of these financial institutions believe that AI will be pivotal to their success in the next few years. So we're seeing companies look to automate critical business processes such as risk management, we talked about fraud prevention. There's also customer experience examples that we're looking at such as chatbots and intelligent recommender systems for retail banks. From where you sit, what are you know, what applications are powered by AI that banks are thinking about today? Are Where do you see AI going in the banking industry?
David Dobson:So I think as our focus for today's conversation is going to be more at the edge, I think I'll double underline that, you know, the security risk management areas is pivotal to the success of banks. And if they're going to loan money to people, they've got to assess that risk. If they're handling people's money, it's got to be secure. So they've been looking at the use of AI in those spaces, mostly at the enterprise level for many years now. And probably most people who are familiar with this area have probably read one or two examples about how banks are leveraging this technology at scale in those kinds of big enterprise workloads. At the edge. You know, we work with a lot of partners around things like ATMs, automated teller machines, kiosks, Digital Signage solutions, all of which are providing either self service type of solutions or information feed solutions or interactive systems that people can, can interact with typically across touch screens or through those tiny little cube wallets where you enter your PIN number, etc. And in that space, I think there's two main areas that we're seeing banks investing. One is, and I think you've mentioned it a little bit chatbots. Now can we bring AI to address a more more natural language way of interacting it, especially in the kiosks or in the self service space, it kind of happens online, now, you're on a site, you start to click a few different pages, and then up pops a little chat bot and says, Hey, can I help you and you can start to have a dialogue in the kiosk world, that dialogue may well be through a voice type interaction. So people are experimenting with that in that self service space. And that's definitely you know, a significant use of AI. And then the other area is actually, it's kind of, it's not as exciting as chatbots, and cool stuff like security and risk assessment. But it's around preventative maintenance. You know, as you're automating a lot of these systems, you want to make sure that they're working, that they're always available for people. And so using AI technology to predict when these devices are going to fail to ensure that you're maintaining them in the right way. You're not there's nothing worse than and you see it less and less these days, but nothing worse than walking into a screen and seeing, you know, blue screen message across it instead of an advert telling you about the latest product that the company is trying to interact with you with. So that preventative maintenance or going to an ATM machine to get some cash and find that it's out of cash or it's got no, it's just not operating today. And it's 10pm on a Saturday evening. So all those frustrations, preventative maintenance, it's got a real role to play in that, not just around that technology, by the way. Also in the in the pure operations. You know, the heating systems or lighting systems in these remote locations, is another area where people are looking at the use of AI to predict and to manage those fixed installations.
Jo Peterson:So we believe it or not, David, it looks like banks and retail might have something in common in that they have, they both have tasks that are not too exciting, but that AI can help with it. You just mentioned that in the banking space. Preventative Maintenance is an important area where AI can help with things and in retail inventory management isn't exciting. But helping with that is also important. It's a known entity and something that is an operational task that needs to be maybe handled better more quickly. And if AI can help there, I'm sure it's, it's going to be a needed piece. But generally, retail is focused on better experiences, right? And if we look at AI right now, in the retail space, at least we're seeing it used in CRM software to automate things like marketing activities, predictive analytics, identifying, you know, what things customers are going to buy? What are some of the other things that we're missing, or that you guys are seeing folks use AI and retail for?
David Dobson:You know, these are such hard questions early in that you can almost say AI can help you almost anywhere in your business. Right. And I know it's kind of a trope, but it probably is true. We and you're right to say retail, and banking and the other industries we'll talk about in hospitality. One of the things we're starting to refer to these industries as internally as consumer industry. So they all directly interact and interface with with consumers in the case of retail, its shoppers. And, as you say, understanding those shoppers, what their behavior is, how best to serve them with products and your services is the most critical thing for a retailer to be successful. So understanding that shopper in whatever mode that the retailer is working in, and you know, this show that there's so many different retailers, whether it's a Walmart that's dealing with 10s of millions of customers every single day, or if it's an LVMH luxury brand type customer where the customers less frequent, but the interactions that you have with those customers can be more important from a store staff perspective. So we talk about things like experiential retail curated retail, and we talk about frictionless retail more with the likes of the Walmarts of the world, and AI has can help and assist in all of those types of interactions in the experiential retail. It can help the store person, guide them through information about the shopper that can help them create a better interaction to create a better prospect of fulfilling the shoppers requirements or needs. So that could be, we've called it clienteling. In the past, it could be solutions that provide information about previous purchases, the stuff that you can get online, that in the store experience or in the past in person experience is more difficult to get ahold of. But that AI technology can really help with those those things. On the friction less, you know, if I just want to go in, I just want to buy my five things and get out well, how can I optimize that store environment so that I can just do that right, I can do it in the easiest way, in my preferred retailer in the way I want to shop with them. Because one of the biggest challenges for retailers is how can they bring people back to the store because that's the most profitable way for them to interact and serve the customer. So there's a lot of use of AI, in terms of understanding that customer behavior in the store, and being able to operate the store in a way that's most efficient for you as a retailer, but also creates the optimal shopping experience for your for your retailers. You mentioned inventory, you've got to have the right inventory on the shelves, it's no point going to your local store and finding your product you've driven all the way down to get or cycled or water to get is not on the shelf. That's the most frustrating thing. So getting your inventory rice is super important. And the supply chain challenges of today, with labor shortages, energy costs, geopolitical issues, and all those other factors that we will read about AI is a great way of looking at how you can optimize those inventory levels and make sure your supply chain efficiencies are minimally impacted by all of these additional pieces. And so, so many different uses. And I know we don't have an awful lot of time today. But if I was to summarize it, I would say in the online world with where retailers, you're interacting with the shopper through a mobile phone or through a website, or some some other kind of app, they know an awful lot about the shopper, and they can curate the experience in a great way. You go into the store environment. Unfortunately, they know very little about that shopper, they may know what you've purchased, they may know how many people visit the store, they may know their inventory levels, but they really find it difficult to create that personalized shopping experience. And I think that's what retailers are looking well, I know, that's what retailers are looking at AI to help improve that overall experience for the shoppers when they visit the stores and environments.
Maribel Lopez:One of the things that's so interesting about the introduction of edge computing in areas such as retail is we, we talked a lot about moving everything to the cloud. But the use cases that she were just discussing, a lot of them require the possibility of real time streaming analytics, which makes much more sense to have processed at the edge. Retailers talk a lot about computer vision and how that might assist in trying to figure out if there are things like stockouts that they need to deal with, or how to redo flows in a store to increase footfall to make certain displays more attractive. And all of those are great examples of how AI and edge come together to deliver from a technology standpoint, what you were just talking about. And I know we've talked about retail, and I know we've talked about financial services, but I don't want to forget about how AI IoT and edge happened in the hospitality industry. I did Jo found a research study that said 62% of customers cited convenience as the number one reason for choosing a restaurant and in a hotel study 78% claim they will only engage with personalized offers. So from this standpoint, maybe we could spend a minute talking about how AI and edge help the hospitality industry differentiate themselves. Yeah.
David Dobson:Just before I dig into that a little bit, I just wanted to underline what you were just saying, you know, I've worked in this industry, these industries, and you were very pleased to just mention to my two most recent companies, but I've worked in this space for 30 years. And the very first interaction I had with a customer. They told me, David, we want to get technology out of our store. And over those 30 years, all I've seen is people put more technology in their store, in their bank, in the hotel in the restaurant, for many of the reasons you've just described, because they want to provide better services. And the way to do that is through that technology at the edge. And with the introduction of computer vision as drive that's driving the need for even more more compute than I've ever seen in my 30 years is required to do those types of workloads which are super important if they're going to really leverage this technology that we're talking about. So coming back to quick service restaurants and hotels. I would say you know In the hospitality space, we think about three different spaces, the quick service restaurant, which is all about convenience and scale of operation. So as you say, is the quick service restaurant close to me? Or will they deliver the products to me quickly? Clearly, if it's a cooked product, it has to arrive, still in shape for me to eat. And look at AI, there's a couple of areas where we're seeing people look at AI at the edge, in terms of use of that technology. One is an area around kitchen safety, more and more people are concerned about how their products are made, what's the position of that of the kitchen, are the staff doing the right thing, and the people who operate those chains really want to make sure that they don't fall foul of any kind of health and safety or any kind of concerns from the the shoppers that the burger or whatever it is that they're cooking for the person is provided to them in the best possible way. So health and safety and kitchen safety is one of those bigger areas for AI to be used. And more and more by the way, that's bleeding into sustainability. So are we reducing waste? Are we making making sure that the, the way we produce and the amount we produce is minimizing that wastage so that we maintain our sustainability issues and goals? And then the other piece is around the drive thru, and the pickup experience or the delivery experience? So the use of AI to optimize routes? And to find people in various locations? Or was that person parked or in the restaurant, they ordered through a self service kiosk? Where are they seated, there's loads of different use cases where AI and computer vision can really help both that the person visiting the restaurant but also at the restaurant, and then move into it, I'll skip on entertainment, because if soon as you start to get into large venues and theme parks, it really moves in towards operational efficiency for the theme park or the the entertainment venue, operator, but also really understanding that customer, which we've talked a little bit about in the retail and the banking space. But understanding the entertainment journey that you're undertaking, and how to optimize that and create the most compelling experiences is really where those guys are focused. And as you say, in the hotel, yes, there's the discount side to it. But I think there's a loyalty side to it as well. I, as a business traveler, and business travel is coming back is super important for the hotel industry, and providing the right level of service to those business as well as the the guest who is on a an entertainment journey as opposed to a business journey, and providing that loyalty and providing the speed of check in speed of checkout. Is the room configured in the right way? Is it in the right location, let's make sure it's it's configured in the right way for that person in advance of them arriving there. All of those things. AI has way is a very useful tool for both discovering at scale, what it is that you need to do, and then optimizing those environments for those people in that area. So covered a lot of ground there. But hopefully, hopefully that was was clear.
Jo Peterson:That was awesome. Well, we, you know, you gave us lots of great things that we're going to be seeing. And in terms of AI and edge and retail banking and hospitality in next year, two, three years. If you had to say one prediction for 2023, what would it be?
David Dobson:Okay, so I think, CES, why, wow, I wish I had a crystal ball, it's not going to be covered 20 It's not going to be COVID 23. So I think what, what we're seeing is, there's a couple of things where, and we've talked about it a lot, you know, understanding more about that customer journey at scale is a key problem that all of these businesses are trying to crack. And so the use of computer vision in solving those problems. And the way to instrument at scale is one of the biggest problems. So if you can't put 10s 15 and 20 100 cameras in these remote locations, over 1000 locations. So what we're seeing is people go into the trial stage where they are instrumenting key locations and trying to learn from that and scale. I think over time, they will be able to deploy large amounts of cameras in these locations and the compute to support those large amount of cameras will become down to a cost point. We're really that will change the nature. I don't think it's going to be 2023 is going to be a lot further out than that. But I will say, I do think that computer vision is the space where people are moving out of experimentation, moving towards small scale deployments to really drive some of that insight and understand what that technology would be. And then from the flip side of it, if you look at what all of these businesses are trying to do, we're using this term, which we've stolen from our friends in manufacturing, is build a digital twin of these environments. I think, once they have that digital twin, once they have a digital representation of these physical locations, that's not just an academic recommend twin, but it's actually an actual twin that's informed by data that's coming from these remote locations, they will be able to move more towards predicting what's going to happen in these locations. And then obviously go on that journey from prediction, and right the way through to automation. And I think that's, that's the other piece that I would say, is top of mind for all of these businesses, there's that kind of digital twin representation of these physical environments.
Jo Peterson:That's pretty cool. Well, we always end the podcast with another tradition. And that's going to be sharing a fun fact. Back Could you share with us?
David Dobson:Wow, okay, why only one money allowed? 1am I, okay. So here's my fun facts. So I tested this on my daughter earlier. And she laughed. So I thought, Okay, that's good. So you may not know this, but Google search, everyone uses Google search. Even as an ex Microsoft guy, I use Google search. If you type askew, as K Ew, in Google search, the page that will come back to you will be tilted very slightly at 15 degrees of normal, which I thought was quite a fun thing to do. So everyone who's listening to the podcast, can go away and try that now. Just type a skew into Google search, and you'll see all the texts go askew. As
Jo Peterson:you know, I'm going to do that as soon as we finish the podcast. That's great. When you can Lovely, thank you so much for all the great insight that you shared. I was kind of taking notes. So thoughtful, and we really appreciate you taking time with us.
David Dobson:Now my pleasure. Really, really great to have conversation and, and great. Great sharing. Thank you very much.
Jo Peterson:Awesome. Thank you.
Maribel Lopez:Thank you, David.
David Dobson:Thank you