SAP’s Bharat Sandhu on AI in the Enterprise

foreign McGuire and on today's espeaks we're talking about artificial intelligence and automation two topics that are deeply intertwined to discuss that I'm joined by Bharat sundew senior vice president application development and AI at sap brought happy Friday to you and thanks for joining us today we're very happy Friday and great to be here James thank you so I obviously you know artificial intelligence has seen an enormous amount of interest and investment in the last several months ever since chat GPT debuted in November of 2022 it almost feels like a semi you know semi-freenzy in terms of AI um I mean of course AI is not really new company's been investing in it for years but now it seems like there's a new extreme you know hyper interest in it what trends do you see moving the interest moving the industry now that there's a sort of hyper frenzy going on yeah you know like I think it's always like fascinating to get attention and it comes in Cycles you might remember back in the day when Siri first came out people were pretty excited about talking to a phone right right right and if you follow like show ice and he would uh tear from Microsoft back in the day you know the one of the things about humans is we're just curious by nature and we want to learn and we love to like communicate and when we have a new breakthrough like an AI we've had a chat GPD which allows us to kind of itch you know like scratch or itch for both curiosity and talk is very fascinating that's where you could have seen so much hype and so much like just people anybody and everybody now talking about AI in a tangible way in the cloud GP right so I think that's what peeling the frenzy it will also I think over time you know come down and go up again uh but and that's normal but the good thing though is it's made AI mainstream like never before AI has always been a topic that's been hard to kind of comprehend before it's like oh yeah I'll say yeah and you know there's some data science teams maybe you use it even though frankly everybody uses it every single day like right you know for Uber for example if you order it right that's using here to match you with the right driver and so forth right so but it just made AI things much more top of Mind conversation uh for everybody recently do you feel you know being in the Enterprise as you do and being so closely to the tied to the sector that there's a sense among companies like oh goodness we know of AI we knew of AI before but now we need to really get on the bandwagon or or as a sense like Wells you know budgets can be allocated one to three years in advance so it's not really quite the the spending frenzy that we might think no look I think every every organization every like and you know uh all the leaders and organizations right now are always always by the way looking to see how they can run the business better and and AI is really forcing them in particularly the really latest friends here on AI is really forcing them to think how fast they can move forward now let's be honest right companies have been doing AI for a long time some more advanced companies have been doing it much better somebody like more mainstream companies are doing AI for a long time the issue has always been is it's been limited to the expertise they can have in-house from data science teams and machine learning Engineers or the ones they can get through the partners and Consultants or other vendors um and and now companies are saying hey can we do it faster because it's maybe easier for more people to actually build around Ai and more importantly will it will this new generation of AI generative AI really transform and impact my business in a big way that I need to be ready for or I take advantage of because nobody wants to be left behind and right now A lot of people are saying hey this is am I going to get a Competitive Edge or I'm going to lose a competitive match so that's really fueling a lot of I would say renewed interest in applying and deploying AI at a much broader scale than before yeah I think you make a really good point in that you know company said been well aware of AI and I think companies have been investing it it's not it's not really new as it seems um but I think companies have sometimes been frustrated with it because it's been expensive and and maybe they're not sure what they want to do and I think in some cases there have been some expensive science experience experiments so to speak it's it's they've tried it and they spent a lot of money and it's like I'm not sure we've got it out of that so what what advice would you give companies to in terms of optimizing their AI strategy or really the AI and automation strategy yeah look I I as much as you as there's interest in hype you know but like in any other uh project investment let's always start with the business outcome in a clear Business Solution set up so there are three things we always look at and work with customers and sap have a unique approach we can come with later on but but basically any customer working with anybody should be looking at three things one is uh instead of looking at technology for the sake of technology that they should do all the time anyway to make sure they're staying competitive but when it comes to applying this person needs to have a clear identification of the business problem that needs to get solved in the potential area that is where a lot of science projects don't even start because they start have a technology let me try to find a use for it versus have a real business problem here or anticipate a real business problem here and maybe AI can help second getting the right set of stakeholders is super important it is not just a data science team it's actually most importantly also the the business team that will be using you know the organization they'll be using and getting stakeholders up front um and and this is an area where people always struggle because the data science teams will always expose explain about models and data drift and all these terms which people have a hard time understanding so you really need the third thing that somebody or a team that can translate business and technology and do the bridge so start with the business it sounds very normal but frankly this is actually the the rigor of this approach is very important what is the business impact business problem we need to solve can a help here then how will we roll out uh test it and roll it out and and any technology adoption whether AI or not the end users will have to use it have to be involved up front not at the end because otherwise you'll have no adoption I think you'll be stuck in an experiment all the time so those will be the the biggest things the one last advice I would just have is uh this is an area where customers have to develop expertise in depth also in they cannot Outsource the knowledge that they should have they should have their own in-depth understanding of AI they should have teams that understand the space because it's another thing that you just want to always Outsource it's inherent capability when I develop just like you develop any other inherent capabilities of running Supply chains so you want to have that in-house point of view all the time also and the reason being you'll be less reliant on a vendor or a partner or a supplier from providing the point of view that of course analysts should be leveraged left and right but the second thing is also nobody knows your business has better as good as you the longer you do it the better you'll get added it's a really important point that companies as as confusing as the technology is and I think AI is more confusing than earlier Technologies it's more complicated than Cloud it's more more complicated the data centers to be sure so I think companies have maybe lagged behind in terms of really having that in-house expertise but I see the importance of it the other point you've made that I think is so interesting is that there needs to be someone there who's a bridge between the technology and the business and I think those people can be rare because there's Executives have a lot of you know business knowledge perhaps and there's there's the I.T folks who really know the technology finding that individual with a bridge uh is is harder to do um I don't know I'm not sure my question is I think it's like how does a company find that person maybe you don't group that person in-house you know it's very funny like the uh back in the day when business intelligence was the you know the right thing right right there was a the there are armies of business analysts that's what they do mostly right that they understand the business problem they understand technology good enough and they're able to translate it here we don't need uh we need really that kind of thinking and that kind of knowledge applied hair um and I think that is the most important thing now a lot of technologies that are making it easier also for the the technical translation to be done as generative AI is already being used to explain some of these things but but I think really the business analyst kind of mindset and that team can be easily adapted yeah and most companies have business their organizations right but there's a little bit of training required there's a little bit of uh required but that I would I would recommend people start there well let's let's drill down to the sap product offering obviously sap has many large Enterprise clients how is sap addressing the AI and automation needs of its customers what does it do in terms of its product line yeah you know sap we have a really straightforward thing right we at the end of the day our business applications company we're an Erp company and we help customers run the supply chain finance and HR and you name it these Mission critical Business Systems so when it comes to AI we have a very clear and straightforward AI approach which is we are a business AI company let me explain what that means um so and I came from Microsoft previously my previous role there we were a generic purpose Azure Azure as a generic platform okay and it's an awesome platform and AWS has an awesome platform gcp has a great platform sure sap what we do is we work with the best AI Technologies in the world we combine that with uh 50 years of data and knowledge in every single industry and customer data and we bake it into our Solutions so customers can access our AI capabilities and our finance supply chain procurement Solutions without having to spin up data science teams they get it from day one it's very relevant for the various Industries whether it's recruiting where you're trying to find Talent whether you're writing job descriptions in a compelling fashion where you're trying to do interviews and you're trying to have audit audit generated interview questions so you make sure you can get the best interview candidates in or actually clothing books faster which is like you know you have multiple organizations they're closing books in a bigger company but they all need to reconcile using AI to match automation or payments automatically and so forth right yeah like literally like uh under the use cases in our capabilities that are out of the box so with sapr approaches business AI which means we give our customers built-in business ad capabilities and the applications they use all out of the box every single day it's relevant for their industry and their business because that's how we train the data and it's built in a highly responsible and trusted fashion and there's always human in the loop so you can be deployed with confidence what we do with AI is basically enable customers to deploy AI at scale without requiring extensive armies or data science teams and solve the most Mission critical business problems interesting okay so sap is is is is um offering the technology because also I'm assuming also offering like a Consulting role as well you know the good thing about our technology is we we make sure our customers don't have to worry about what AI model needs to be used we take care of that in our applications so it's really out of the box when customers get access to that so even you don't need Consulting teams also by the way now we of course provide business Consulting that's because our customers we work with them on literally running their Finance operations and the supply chain so there's a lot of like business Consulting that happens with that that's part of that um we do that but you don't need and of course we have technical teams that can help with custom AI deployment also but our goal actually I don't know what end of the day is when for instance uh somebody's deploying a supply chain solution and and they have trucks rolling in every day with you know widgets coming into the factory right the documents that the truck delivery it says that I delivered these thousand widgets not right now people have to manually enter it then what we've done is automated the digitize all that thing automatically you scan a document it goes in automatically in any language in any template if you use generative ad for this and then we take it out of the loop so no technical is nothing required it's literally part of the supply chain solution ah interesting yeah and that's that's actually what I find really fascinating when I connect with customers because you know what we're not talking about deep AI concepts with them we're talking about business problems and say hey this is all built in right and where we need to do custom we can do custom but that you know like with any custom work takes a little bit longer interesting is is there one I don't know if this question can be answered if not we can we can skip it but I think the interesting question is is there a more typical use for AI among those large Enterprise clients is like typically this is the default thing that they do first with AI or not necessarily it's gonna it's gonna vary so much uh no it's very varied but I'll tell you the most common usage patterns we see so one is like uh in finance it's really helping you use AI to close books faster right what do accounting teams do basically in a bigger company they initially say I have so many invoices I have all these uh iOS and supplier codes I need to match all the stuff up that's a really great machine learning problem because I have a bunch of invoices coming in emails I need to match them and close them out and everything else right Swiss real is a company they process a million payments of you know I think every year or even faster they match 99 of them using AI um you know and nobody manually has to go in and have to say Okay this thing matches here I need to extract this price point or this payment and put it in this line item here we do it automatically right the supply chain and procurement let me say when when customers are looking a very huge business Network which basically is a network of suppliers and customers together so when customers are looking to say I need to uh buy a new widget I'm looking for a new supplier you know we help customers say what are you looking for and then we use AI to recommend suppliers and then when they're building an RFI they say these are the questions you might want to ask right so in procurement and then once customers are sending start working with the supplier onboarding suppliers all AI assisted also so those are the other areas right the other thing we announced recently was the sap digital assistant which is like say I want to create a sales order from a PDF how do I do it and it does it for you does it helps you walk you through it right but helping engage more from a ux perspective with our business applications is a very very common way that our customers are doing it but literally you know it's uh it's depends on the department and the most common use cases that's what we put air to use oftentimes it's transparent customers don't even know here is being used right right but it is helping it's streamlining processes yeah um I think the big question that companies are curious about is what is the future of AI and Automation in the Enterprise I mean inside yeah it's a very open-ended question it seems like with generative AI I think it could go in any number of directions yeah what do you predict over the next oh a couple years or or so look I'll start with a very grounded reality where we are right now uh and then I was more greater and more like where we hope I hope we can get to as a as the industry okay today most customers have silos of disconnected systems um and you know I never know what no one company will run on one vendor ever and you know this should the different Technologies for different things but for complete automation uh you know you need to have a very connected ecosystem so first I think a step right now which the customers are doing really well already which is more in higher maturity is integrating their various systems and Business Systems together right using whatever solutions that sap solution or integration suite for that but that is where the higher maturity is right now so that's good the second thing is automating processes the span multiple systems right because then you cannot use just embedded Ai and one solution you have to use literally like connect the different systems and automate them across so there is a concept of Enterprise automation which is really using process insights and using AI to really understand where the process bottlenecks happening that touch multiple systems and not just doing it up front but actually doing it all the time so if you have an amazing Optical signal we had to enables customers to do process mining in a continuous process then based on those insights you want to trigger automations all the time for example Give an example at the end of the quarter lesson in the company the sales there's a ton of sales order that need to be process the discounting rules need to be relaxed right and instead of getting a manual check every single time in this you automate it all the time but we can predict how we can actually measure if the sales approval backlog is increasing and if that happens at the end of the quarter some conditions are met we should automatically approve these things right so creating these Enterprise automation systems is the next step now that we've integrated these things I see that happening and over time AI also then starts working across multiple systems in an automated fashion so that's where I already see the evolution of an AI based automation system happening which is it is more proactive and is more proactive across the customer's unique needs and it truly becomes an assistant and helps customers augment their business systems so artificial intelligence becomes this uh sort of super intelligence that combines the the Enterprise solos and get the silos and it gets them all working in concert among other things yeah I I don't know I won't go to super intelligent but it's just like more tuned to every business specific uh characteristics like a manufacturing and child manufacturing channel would be pretty different than a retail operator in in Europe um and but you know and and then would it not be nice as AI technology involves that they connect all these systems as because we already are able to do that but then AI then becomes much more fine-tuned to each of these Solutions Industries um and there's not a lot of work that needs to be like there's a super far out but think of a personal digital assistant for the business uh-huh I guess that's the key phrase the personal digital assistant for the business yeah that's right that's right like like if you look at all the Sci-Fi movies right the personal digital assistant that works really well for us personally is the one that knows right the ones that will work best for the business are the one that know the business systems the best and that's frankly going to be investing uh as a company at sap right that's fascinating uh it's a lot of good stuff and I think it's going to be fascinating a fascinating sector to show you watch in the years ahead and certainly the Enterprise uh users are flocking artificial intelligence uh thank you so much for for sharing your expertise today and I hope you come back and talk with us again sometime yeah like wisdom thank you so much foreign

This transcript was generated automatically from the video's captions and may contain errors.