SCB Data X, XL Axiata and Dell Technologies: Navigating the Generative AI Landscape

In this episode, we sit down with  - Dr. Shuki Idan, Chief Scientist, SCB Data X, Muhammad Sufyan Masood, Head of Analytics Center of Excellence, XL Axiata and Richard Jeremiah General Manager Data Centre Solutions, South Asia & Emerging Markets, Dell Technologies- to explore the impact of generative AI on business models and operations.

From driving productivity in sales and customer service, to optimising critical infrastructure investments, these experts share their first-hand experiences in leveraging the power of AI to gain a competitive edge.
They discuss the challenges of responsible AI adoption, including data privacy, governance, and change management - providing a roadmap for organizations looking to accelerate their generative AI initiatives.

Whether you're a business leader, technology executive, or simply curious about the future of AI, this episode offers a unique, insider's perspective on navigating the generative AI revolution. 

Transcript

0:00

Across Southeast Asia, enterprises are re imagining what's possible with artificial intelligence. The conversation has shifted from if to how. For the first 20 years, I had to convince the organization to use AI, and this was a tough task, you know, to come and say, We can do it with AI. And nobody believed and asked to get the proof, to get a POC, to get some some demonstration. And today I, I find myself in in the opposite sometimes direction when people say, let's do it with AI. Let's do it with AI. You know, like you have a hammer in your hand and everything looks like a nail. And sometimes I say, No, no, no, don't. Don't use AI here. There are other ways to do it. Beneath the excitement and promise lies a sobering truth. AI is only as good as the foundation is built upon. Data is the fuel for AI, and you know, at the end of the day, bad data, bad AI. So at the end of the day, we feel, well, most of our customers data security is critical for every organization to protect their IP. For those who get it right, the rewards are transformative. AI isn't just a tool anymore. It's becoming the core of business strategy. For the key thing is to focus on the impact. I think, because once people start looking something you know, tangible, they get attracted to it. And the impact with the we can say, in a short time period. So and Gen AI gives us that flexibility, that we can produce the stuff very quickly for the people journey to AI transformation is complex, but the path to success lies in understanding both its potential and pitfalls from Southeast Asia's leading enterprises. This is a special episode from the Dell tech forum. Welcome to the AI BP ASEAN B to B growth podcast. The aibp ASEAN beach to be growth podcast is a series of fireside chats with business leaders in Southeast Asia focused on growth in the region. Topics discussed and include business strategy, sales and marketing, enterprise, technology and innovation. You a very warm welcome to all the attendees joining us at the Dell tech forum today. Well, my name is Wai, wai, and I'm from aibp, and we serve as avenue for public and private organizations in Southeast Asia to access and exchange information about growth and innovation within the B to B space. We have a current network of over 30,000 stakeholders in Southeast Asia, and we develop ecosystems by engaging in activities with partners such as Dell to create value added information for our stakeholders. Well, today, we're here to discuss about generative AI and applications within the enterprise space. Very interestingly, we've been tracking how enterprises in Southeast Asia are really adopting AI generative AI within the organization in our annual enterprise ASEAN awards, Innovation Awards, which seeks to recognize enterprises seeking to make transformative impacts within their business through adoption of innovative technology, we have seen 72% of all applications talking about AI in Really adoption within their organization, driving innovation for their business. So today I have with me Doctor Shuki Idan, the chief scientist of SCB data x pilot, together with Muhammad Sophia Masood, the head of analytics Center of Excellence from XL xiata Indonesia, and Richard Jeremiah, the general manager of data center and compute solutions from South Asia, Dell Technologies, gentlemen, welcome. Well, we have heard you and your organization. Perhaps I would like to start off today's discussion by having you share a little bit more about your own journey within generative AI, shall we start with you? Dr Shuki, yes, I can do that. So my name is Shuki done, and I I've been in this space, I think, 30 plus years. So, you know, I can say that for the. First 20 years, I had to convince the organization to use AI, and this was a tough task, you know, to come and say, We can do it in AI. And nobody believed in us to get the proof, to get a POC, to get some some demonstration. And today I find myself in the opposite, sometimes direction. When people say, let's do it with AI. Let's do it with AI. You know, like you have a hammer in your hand and everything looks like a nail. And sometimes I say, No, no, no, don't. Don't use AI. Here. There are other ways to do it. So I think that the wonderful things happened when I started my journey there was there was no data to learn from. We didn't have the cloud, we didn't have compute power, we didn't have open source and for sure, there were very few that were in this domain, and now it's in the hands of everyone. So today, we are going to speak about the enterprise. But what happened is that each one of us is using this I think it's a fantastic breakthrough, and so everyone can benefit from from this technology. But every good comes with some caveats and some red alerts. You know, every technology that becomes mainstream is first adopted by the bad guys. So people are doing the deep fake. People are blackmailing things like that. Security becomes an issue. Uh, bias, ethics, all kinds of things become an issue. And I think we will discuss it during these discussions. Thank you very much. Doctor Shuki, we talk about the AI winter last time, 20 years ago. Now we are entering the AI spring right where everybody is talking about Gen AI, you are definitely right with regards to the challenges or the flip side of AI, which will go into it a little bit more during our discussion, but perhaps I'd like to shine the spotlight on Sufian. Suf share with us a little bit more about your journey in Gen. AI, well, yeah, Sophia, I'm here, and I actually started my journey around 18 years before in business intelligence and analysis domain, you know, AI wasn't there at that time, or was somewhere in the dreams. And I have seen that evolution of data and analytics, starting from very simple databases to big data, and then moving it, taking it to the cloud environments, and then AI came into the picture that how, after making those data lakes, we can start using unstructured data and start building things on top of it. Now in just like Dr Suki said, it is a very interesting situation that where we have to sometimes stop the people that for everything you don't need AI so no need to put everything under the umbrella of the AI. There are many things that can be done without that, but of course, now it is showing wonders, and there are many, many things which we are able to do with respect to getting or generating the new revenue streams and coming up, introducing the efficiencies in the existing processes, enhancing our customer experience. And you name it, you name it, there are many things which can be done now. We started setting up this AI setups around five to six years ago, when I was in Vodafone, and in My Vodafone career, when we were introduced, I was actually introduced with AI and we started setting up and into the IoT platforms with respect to different commercial use cases of IoT and how we can introduce AI over there to enhance efficiencies. And since in Qatar, FIFA was coming over, so there was a good chance and investments were there as well. So there were good chance to introduce couple of solutions, but slowly and gradually, we saw that there was a boom in the AI, and especially after Gen AI. Now again, we have to kind of educate people that after Gen AI, it doesn't mean that you're going to put everything on the Gen AI. There is still classical AI. Into the space. So we have to see that what what you have to do with, what technology in a cost efficient way. So I think this is where we are. And I see definitely a lot of things on the road and that that can be very beneficial for the overall societies. Well, Sophia, thank you very much for your sharing what we've talked about, FIFA and the very nice link to technology. Well, Dell is also a proud sponsor of McLaren, whom in the recent Singapore f1 race was the proud winner of f1 here well. Richard beyond your role at Dell, tell us more about your personal journey of how Gen AI has touched your lives and in your 16 years at Dell, how has that changed? Thank you. WABA, it's absolutely my pleasure to be on this stage, and, you know, with these esteemed guests as well. So thank you, Dr Sophia Suki and Sufian as well for being as part of this. I'm really excited. You know, 1956 is when, you know it was aI was started in a university in the US. So it's been there around for some time, but Gen AI is really got this into a rocket ship over the last 1824, months, or maybe more. I mean, my experience with AI has been, you know, if I look at a use case that we at Dell, we found out our sales people use about 40% of their time trying to figure out for their sales preparing for their sales meetings. You know, if I'm going to go and meet Doctor Suki, we spend 40% of our time trying to figure out where, what is done, who is involved with all that side. It takes a lot of time preparing for these meetings. We have involved Gen AI and into that sales process, which it'll go and figure out, you know, from social media, from everywhere, to figure out where and what. And in a one pager, it'll give the sales guys, you know, you know, at the fingertip, information about Doctor sake. Thank you very much, Richard. If we look at how the ways that Gen AI has impacted enterprises and the business model, how have you seen it really transforming the business, perhaps you can start off with yourself, Doctor Shuki, you have seen the growth of Gen AI in your various roles, from technology firms to Agoda to right now, really driving a new business at SCB. Tell us more about how Gen AI has impacted the business model and the organizations you've been part of. So I think that until now, it's not, I didn't see really a end to end, or a big transformation from the user AI. It's, if I look at the past, you know, before Gen AI, it's mainly focusing on some kind of metric or improvement you want to achieve. We want to achieve more sales. We need to identify the best leads. We want to improve our customer satisfaction so we process information of feedback. What's happening now is that we are trying to broaden the scope of AI so we are able to use it in, let's say, in a more massive way, to change the business. The changes that we made was that we believe in data and AI. So everyone should think in this way. Everyone should measure. Everyone should, you know, even access data, which is a small part of AI, is super important, super important for people to be able to look at data. And I'm not talking just about the CEO or the general manager. It should be in the hands of everyone. So there are all kinds of Horizon, original concepts and change management that happened that are even before the big the big wins. And the other side of it is that there is still a risk that not many can digest, especially in large enterprises, and in my case, also regulated enterprise, that the technology comes with the risks, we have to remember that AI is a statistical solution. Sometimes it makes mistakes, so in many cases where small mistakes can have a larger impact, we are still reluctant to to implement it. So you. I believe it's a journey. You know, we are checking compliance today. Of the things that are said on calls with AI, we are supporting our agents on investments and on handling customer requests. So in most cases, we have still people in the loop, and it's not really our strategy as a business did not change, but the most important thing is that we it became a driver of our strategy. And I think this is the largest thing that happens, or it will take some time to settle and to coordinate the organization from top to bottom and from bottom to up. So I believe, for large organizations, still, we don't see major changes in strategies. But again, the change in strategy is the mindset, is the understanding that the data has value, that we can learn from data and do smart things, even things that we don't understand how they operate. You know, no one can. Of us can say, you know, what's in the mind of others, but AI can at least tell you what it what it's in the mind of the average other. And like any muscle, right? The more you use it, the stronger it becomes. And with such a powerful tool, of course, there is a lot of responsibility in using it ethically, and what you've mentioned, productivity has definitely improved with Gen AI applications within the organization. If I could go to you, Sufyan, one of the interesting case studies that we understand your center of excellence has led is really looking into AI driven capex prioritize prioritization. So I think for telcos, with the event of 3g 4g 5g and now 6g there's a lot of investments that has been poured into new investments into infrastructure itself. And tell us a little bit more about how you have used AI to help prioritize CapEx spend. Sure. So in the telecommunication The one thing is the fact it's a very, very dynamic industry, and we can see around the world, consolidations are happening, and if we see that, usually, the prices of commodities, we have seen, even due to the inflation, has increased. But despite the fact the telecoms products are still, you know, the prices are decreasing. So, so our margins are getting squeezed. On the other hand, sites, we have high demands of services, high end services, the video streaming, high end video streaming, and people are asking for high end experiences. So we have to focus a lot in terms of our smart investments, that wherever we are going to give our network that should be, uh, wherever we're going to put our web, our towers, or expand our coverage that should be addressed very smartly, and which of the towers we're going to upgrade, and where we need 5g and where we don't need 5g so all those decisions are very, very critical. Now, if I just see from Ai perspective, around three to four years ago, AI was in more exploratory phase, where we were trying to do the things that were trying to explore different use cases and trying to see the impact. But now things have changed. I have seen that AI is really making an impact, at least in the telecom world. And the key thing is to focus on the pain areas that how do we select our AI tracker or AI roadmap? How we bring in the most relevant use cases into the picture, which can bring us the impact so with respect to that, and that can be fulfilled through classical AI or Gen AI, that that's the second part. But the key thing is first to identify the pain areas. Then another key consideration is how we can enhance the efficiency of our sales team. We have huge sales sales teams where this different distribution layers and meeting to the retailers which are across throughout the country. So we usually call it a VA. Have a very known KPI which shows which tells us about the availability of our brand and inventory. Visit. Ability of our brand and advocacy of our Ro So how we make sure that our is intact? Then, then the third thing, another key pain area. The third pain area is with respect to the operation, like the efficiency or operationalization of our services in it and in the network side, how our customer experience is getting impacted with respect to those customer services and how we can make them, you know, more smooth and streamline, then internal company processes where our teams are working and they need analytics on day in, day out basis. So how we can, you know, provide them those required data sets in the required format where they can quickly take the decisions. So these are basically high level pain areas, which I'm mentioning then. So these are all small problems that we need to make sure. Then for each problem, we have to see that what AI can bring in. This is where our use cases start pouring in. And then for those use cases, then we start looking into the impact. This is all the planning part that if we will come up with this impact, how much I would be able to reduce would that be? You know, that would be, in terms of ROI, it will be better, or whatever I'm doing right now will be better, right so after coming up with the impact, then we prioritize our use cases, and we come up with an overall strategy, right? So this is how we select the use cases which make sure that in the end, this impact will be realized. And this is something we not only align with the stakeholders who are going to use that, but also we align it with the relevant finance teams who are going to book the value, and then they have to show that what is going to be the impact in terms of either enhanced revenues or cost cuttings or cost avoidance or revenue loss avoidance in those technologies. This is how Gen AI has really boosted our analytical paradigm and our pricing teams, they are taking decisions on the basis of exact right market information, and then the level of the retailer, their knowledge is the Key thing, because a retailer is approached by multiple operators, multiple competitors, so the knowledge of my product should be in top of his head. Then he will be able to sell my product. Yeah, so with respect to this is where Gen AI, you know, brought in a very interesting change that we have introduced different virtual agent, a sales virtual agent for them. So whatever question they have about our product or something they can ask, so on that basis. So, so this is how, basically, the Gen AI is enriching to solve multiple problems, which are then solving my overall pain area, and in the end, it is sure giving me the impact which I'm able to show to the finance and to the company. Yeah. So this is one area sales. So similarly, in the network site, very big challenge is with respect to the outages which are happening in the network. So how autonomous, uh, networks we can create, which can, you know, identify, or at least predict that where the outage is going to happen? By having a look onto the logs of different systems at the back end, and once the outage has happened, then how we can help the agents to quickly solve the problem? This is where Gen AI, you know, comes into the picture. So we have the Knowledge Hub, which is having all the issues or outages that happened before, and what was the RCA? What were the actions which were taken? So, what are the top key things which should be looked into if this type of outage has happened, you know? So whenever there is an outage, so it starts giving some, you know, recommendations to the troubleshooting people that you have to look into, and we have seen clear improvement into the recovery timings, which usually the outages, which are taking 10 to 12 hours, now they are able to fix in two to four hours. That's very good return or outcomes that you have drive. Thank you very much. Sufyan for sharing that. You know, we just, we had Doctor sugi and Sophia really give us very deep insights about how they're applying AI and Gen AI within their organizations. On your perspective, where do you see the most significant capabilities, and perhaps also, since the two panelists have also spoke about limitations, right? The limitations of Gen AI today. I mean, thank you. That was really a detailed information from Sophia, and I'm really very excited, as you know Doctor Suki and Sophia and mentioned, Gen AI is in sales and customer satisfaction. Voice of the Customer, you know, you're looking at how we can increase productivity customers, basically right across education. So from a virtual co worker, digital assistant, you know, everything is better, faster and easier, and logistics, even you look at our organization, you know, the significant capabilities that it's creating in sales. I talked about right at the beginning. It's, you know, changing our logistics. It's creating a much better efficiency from a supply chain we talked about, so talked about, you know, preemptive, predictive. We in our support as well. When we look at our technology. We've got aI built in every bit of our technology. So when we look at the logs, we look at the patterns, we can straight away predict that this is going to fail, this at this time, and we go and do something well ahead before it fails. So these are some of the things that we are really helping our customers manage some of these challenges and and, yes, of course, there are limitations, but at the end of the day, it's how you manage it within your organization and use it for the betterment of productivity. And, you know, gain that competitive edge, because if you're not in it, you know you're going to be left behind. Thank you very much, Richard, we talked about the applications of how Doctor Shuki and Soufiane has applied Gen AI. But if I could go back, take a start by about how you can set the AI strategy for the organization Doctor Shuki, in your experience, can you also use AI for strategy setting. So I think that today's strategy is mainly decided by humans, but once we decided what, what we do, the good thing in AI, it never gets tired, so, but what we are using, we are doing today is once we have an idea about what we want to do, we want to reduce price, we want to promote some some products in order to go to new markets. We can let the AI tie all kinds of variation of action plans of alternatives, and we can do it 1000 times, sometimes with the same assumption, sometimes with different assumptions, and we will get the 1000s of possibilities and success factors. So even if it's not the strategy itself, the execution and the decision how to do it can be greatly done by AI, accelerated by AI, thank you very much for sharing that. Dr Shuki, well, we've started a little bit on this at the start of today's discussion, right? Really, about how do you manage the obstacles or the challenges with Gen AI Sufian, in your application, you've walked us through quite a number of the different use cases. What's some of the biggest obstacles that you see in really driving or adoption of Gen AI, yeah, there are multiple challenges. And of course, one of the key thing is with respect to the governance and that when we see in our overall, overall AI governance, one of the key component is the accuracy and reliability that this is something which is still, you know, there are some question marks on it. And when we are we develop our use cases as well. We try to keep on checking and that. But the actual thing is that we don't know how the result has been prepared. You know now on what basis this decision has been taken. So this reliability and accuracy thing is one of the key thing. And secondly, is the data privacy. Many models are not here. LLM models are not within, like the borders, so they are somewhere outside in the world, and you are not sure that how the data is and according to basically the. Guidelines, or the privacy laws or our internal guidelines, it's not allowed for us to, you know, send our customer data which is a sensitive data. So, so we have to be very, very decisive or very cautious that what type of data we are going to expose to the models which are not here. And third key challenge is, of course, with respect to the cost, right? Because it's not free, so like currently, if we are, there is a big discussion with respect to automating or putting up this virtual agent at our touch points, where all of our customer interactions should be done through virtual agents. But cheap that will Yes, somehow it will increase, definitely the experience, but it will increase on the costing side as well. So that solutioning should be devised very intelligently, that it doesn't give us Bell shocks or spikes. So these are basically the challenges which we are still going through, or it's it's a journey, and on every use case, we make sure that all those Mayas or all those aspects we are looking into in detail, and then accordingly, then we decide that how to take next step. Thank you. Sophia Bucha, I think what Sofia has mentioned must strike very close to ho right, because sometimes when you work with organizations to really become Gen AI ready. They still cannot risk really forgoing the customer trust at the end of the day. They have worked very hard to gain that. How do they create this architecture that helps to protect the data? Can you share with us a little bit more about how you helped your customers to do that in a foundational way. I mean, we all, we always say data is the fuel for AI. And, you know, at the end of the day, bad data, bad AI. So at the end of the day, we feel, well, most of our customers data security is critical for every organization to protect their IP. So it is very important that they have very clear, you know, understanding of the data classification make sure employees are trained. They prioritize the training data encryption endpoint is where all the data is, is, is really input, you know, in the physical world. So make sure have a very clear end point security system to protect their data and the multi factor authentication. And nowadays we have remote workers make sure we have solid policies, and if you follow these practices, your data will be in the right place, so that you can get the right insights at the end. So these are some of the things that we help our customers with. Thank you, Richard, well, in the interest of time, I think we have many different enterprises organizations joining us today who may be at different levels of their Gen. Ai journey. What are some best practices that you would recommend to accelerate outcomes from generative AI initiatives? I'll go around to all of you, perhaps, think about your journey in Gen. AI, what are some tips or tricks that you can share with all the organizations joining us today about how they can accelerate outcomes from their generative AI, initiatives, we'll start with you. Doctor Shuki, so if I take a step back, on one hand, it's a big change for the organization. And how can we bring everyone on on this, on this journey? So you know, having translating the use of technology to quick wins. Not, you know, just using AI is a buzzword, but at the same time making making it enterprise wide. So we have, for example, a must do training every person in our 30,000 people organization you know, from the teller to the CEO, have to do AI training and get a certification. Some of us can do it quickly, just by going to exam. Some of our need to listen to the courses and then do the exam so can spa. And see about what the other teams did. A lot of, you know, events we are doing buttons, buttons between division, you know, on AI. So we give we do competitions, we go and participate in competition, outside the global competition. Because we need also to learn. You know, when you go to to COVID or to nips, and they have competition on predicting something, you know, sometimes, you know, recent one was about privacy, how you detect privacy issues in with llms, and it has a meaning for our business. So we go and participate, hopefully winning, but for sure, learning from it. And I think for us, you know, beyond what we talked and I don't want to take the ground for other things, but the change management is, is a big thing, and including the risk management of things, the security, the but really, for me, the people are the key, and we have to remember that typically, AI projects never end. So we always collect more data, and we always need to improve things and to optimize so the success is not at a certain moment in time and it's not one number. Thank you very much. Dr Shuki, so far on your side, any recommendations on what organizations should do when starting on their Gen AI journey from ice the just like I mentioned before, the key thing is to focus on the impact. I think, because once people start looking something, you know, tangible, they get attracted to it and the impact with the we can say, in a short time period so and Gen AI gives us that flexibility, that we can produce the stuff very quickly for the people, which can at least enhance their day to day lives. They can, you know, reduce their work. And they can feel that, yes, it has it is bringing some impact. And for that, I think the key thing is how we set up our architecture, how we set up our AI factory, that it is scalable and it is approachable by all the people, or all the business champions in different departments, and now it's not like one team who is doing AI for all the key thing is, how you make it available For all the departments that they can build their AI by using your secure, scalable infrastructure, and then they can see the benefits. And yes, when there is a productionization, then at the back end, you are controlling it, but let them play. Let them do some experimentation. But of course, controlled in a controlled environment. And when we really see there is our benefits, which people want to productionize something, then you know, we have to do it in a proper way. And of course, culture is is the key thing. How would we need to conduct the road shows to the people, and if one department has done something, so we have to show it to the other that this is how it has been achieved. For example, here in our legal department, they came up with a very nice Gen AI based solution with with the help of AI team, which was giving very quick insights of the contract, especially now people keeps on getting lots and lots of proposals from different vendors, and they have to go through in detail. And secondly, if some conflict happens, then they have to see quickly that what is the clause which is in favor of them, you know? So something like this. It was a very interesting tool, which when they built so then they started showcasing it, along with AI team to other departments, which put idea for HR people that, yes, why not? We can use it with respect to the the profiling, or the lots of documentation, which is there for the people who can build something. So this is how you keep showing your success stories. You keep showing the impact around the company. And then you, you know. Uh, keep attracting different people and the once they will be attracted, then you would be having a secure environment over there which and scalable and approachable for them, in the dev environment, where they can play around. And slowly and gradually, you will see there are many use cases around which are giving some value to different departments. Thank you, Sophia and Richard. So could you bring us back home? Bring upon all the different points? I think Sophia mentioned about scalable architecture, perhaps share with us what, how should organizations prepare for the AI or the Gen AI journey, sure. So very, very great points. Sophia mentioned, this is how we are helping our customers accelerate the adoption of AI. I'm going to give it, give you basically a lens, you know, from the technological lens perspective, and so beyond mentioned, we call it the Dell AI factory. So it starts with data. Data is the differentiator. So as I said earlier, you need to make sure your data is ready and make sure it's secure and all that sort of stuff. So number two, it is, you got to make sure you bring AI to the data. You know, we've got 83% is, is on prem, 50% of data, or is, is at the edge. So we got to make sure we take AI to the data, not the other way, not the other way around, and then the and the number three, one size does not fit all so we gotta make sure it's the right size for the right solution. Number four, we have an open and modular tech architecture. So we gotta make sure that, you know, technology is open and so that we can really make sure innovation continues to be happening. Innovation is so quick now, so you have to have an open module architecture number five, basically it's to be, has to be an open ecosystem as well. You know, it's not just one window. It needs to be a partnership with a lot of AI ecosystem brought together to help our customers. That whole open ecosystem, is what we help our customers with, making sure we bring that whole thing together. So we, these are our core beliefs. And, you know, at the end of the day, it's all wrapped up, it's sustainability and security as well. And these are our business imperatives as well. So really, our core beliefs is how we are supporting our customers through our from our technology lens, and that's how we are helping our customers to accelerate innovation and accelerate the adoption of AI. Thank you, Richard, thank you very much, Dr Shuki and Sufyan for sharing your insights. I hope for all of you joining us today, you have some key pointers you can bring back to your organizational leaders about how you can help navigate the generative AI landscape. Thank you very much for your time today. Thank you. We hope you've enjoyed the episode. For more information about business growth in the ASEAN region, please visit our website@www.aibp.sg and.