RISE AI, Part 2: AI Powered Enterprise Reinvention

Episode Topic: AI Powered Enterprise Reinvention

How does a major enterprise move AI from isolated projects to the core of its business strategy? Arnab Chakraborty, Co-CEO of the Telstra-Accenture Data & AI Joint Venture and Chief Responsible AI Officer at Accenture, and Dayle Stevens, Co-CEO of the Telstra-Accenture Data & AI Joint Venture and Executive for Data & AI at Telstra, detail their landmark joint venture to accelerate Telstra’s AI reinvention. Discover their strategic roadmap for fusing technology, talent, and business value.

 Featured Speakers:

• Arnab Chakraborty, Accenture
• Dayle Stevens, Telstra

Read this episode's recap over on the University of Notre Dame's open online learning community platform, ThinkND: https://go.nd.edu/b1735d.

This podcast is a part of the ThinkND Series titled RISE AI.

Thanks for listening! The ThinkND Podcast is brought to you by ThinkND, the University of Notre Dame's online learning community. We connect you with videos, podcasts, articles, courses, and other resources to inspire minds and spark conversations on topics that matter to you — everything from faith and politics, to science, technology, and your career.

• Learn more about ThinkND and register for upcoming live events at think.nd.edu.
• Join our LinkedIn community for updates, episode clips, and more.

Transcript

Welcome and Introduction

1 0:06

Good morning everyone. How are we doing today? Excellent. As I said yesterday, enjoy the two sunny days. That ain't gonna last too long. And welcome to the morning. so hey, thank you again for joining us, uh, over the, uh, yesterday afternoon and even to the morning. Uh, and it's gonna be a glorious day ahead. Forget the weather, don't look at the weather outside. Just look at each other and we'll find joy, and, and when our conversations. So it's my pleasure to kickstart the morning with our opening keynote by, uh, Ana Van Dale, who have flown all the way from Australia, uh, to join us and accompany us, uh, today and, and, and, and share, uh, their wisdom about what they're thinking about data and ai. I'd like to introduce an first who serves as the chief responsible AI officer at Accenture's, guides the company's global strategy strategy on ethical AI over 25 years of experience. Then part patent in, machine learning solutions. He has led. Data-driven transformation for 1400 companies, and also has been a regular critical voice in the global conversation, speaking at the us speaking to the US Senate AI Insight Forum and the World Economic Forum under the profound implications of ai. and he's also the joint venture between Accenture and Telstra leading, co-leading that joint venture with, uh, with Dale as well. And I've known an of now for a couple of years. And even the first time when he came to Notre Dame, we hit off and he, he's, he's a man of ideas and, and not just ideas that what I found out about him as is quick to execution, right? He just builds on something and says, okay, let's go ahead, execute it. So thank you enough for, for joining us today. And, and Dale, thank you. Thank you so much for your first time on our campus. Very much welcome here. And she has had a distinguished 30 year career in technology at Telstra. Uh, she's currently with Telstra and also co-leading the, uh, Telstra, Accenture, uh, On data AI alongside a a, uh, and she's the executive for data and AI at Telstra. Uh, she's recognized as a leader who loves solving big problems and enabling potential. previously serving as the Chief Data Officer at a GL Energy and a divisional CIO at NAB, she's a recipient of an Order of Australia medal for her services to it and to women. Uh, she has championed diversity in STEM and serves as a non-executive director at Beyond Blue. Be in the midst of celebrity. Right? How many of you can say that you've been in a room with an aust with an Order of Australia medal, award? so with that, please welcome Alana Bendale. Thank you.

Speaker 3 2:50

Alright. Good morning everybody. How everyone doing? Awesome. No thanks. Thanks again for, you know, spending time with Dale and myself and really big kudos to, you know, nthe to you and your team, and Notre Dame and Lucy Institute for, you know, inviting us to be part of this. And I genuinely feel that, you know, first time I came, it felt home for me and it feels home for me today. And it is really great to be part of, part of the Notre Dame family. Uh, I think what Dale and I are going to do today is, you know, share a little bit about our journey. Um, you know, what we are seeing, what we are learning, and then, you know, hope to get a good interaction, you know, uh, with all of you in the, over the next hour. So with that, uh, why don't we dive in? You know, I wanted to share a little bit of perspective of what we are seeing around the world, you know, when it comes to all things ai, and then we'll go and zoom in into what we're doing together at Telstra. So if you, if you, you know, look at the conversation we started yesterday evening, and, you know, even that's happening globally. There's a lot of optimism, you know, around ai and it's a, it's a massive tech revolution that's happening. Tech led revolution that's happening. you know, there is a, there's so much of. you know, I would say investment that is coming in, uh, the economic value that AI is, you know, potentially going to demonstrate over the next 15 years is over$10 trillion. It's massive. Right? And, and then there are a couple of things. You know, this is not the first time we are seeing a tech led revolution. You know, we have seen the, the computing revolution happening in, you know, with the microchips in the seventies. You know, we saw the revolution that happened with internet in the, in the, in the nineties and then with the cloud in, in the, in the 2000 and 2010 onwards. But this is different, you know, the AI revolution that we are seeing, especially with the advent, you know, of uh, GPTs and in everything with gen AI is very different. I think it's first time I would say that AI is able to demonstrate that we have gotta know how to handle the complexity around language. You know, machines are today able to understand the language, you know, is able to understand the intent behind that, understand the context, and then be generative and creative at the same time. And the ability to go multimodal along with that is, uh, is, is kind of a kicker on top of it. And, and now with the agent ai, it's taking it further saying that, how can we now extend the capabilities of LM LLMs and put it into action? So I think that's kind of the, distinction. And I think a big distinction around the, this new age of AI is the speed and the velocity. You know, I think all of us saw how when, you know, gene AI landed, you know, late 2021, early 2022, with the Frontier models coming in. We saw how, what happened, you know, with chat GPT, you know, and, and, uh, number of active users, you know, going to a hundred million in, in just a month. You know, that's, that's, uh, an unfathomable. obviously it was, you know, followed with a lot of the open source, uh, you know, capabilities coming in. We are going to hear from meta later today and, you know, we saw what happened with Llama and the Siles of the world, you know, trying to get, uh, democratize LLMs, you know, and, and customize and, and at lower costs. we are seeing the rise of agents, you know, agent T ai, you know, is, is the talk of the town. Everybody's talking about it. No conversation is complete without that. Um. Multimodal is a big topic. You know, we are seeing multimodal AI models coming in. Uh, we are seeing embedding of those AI models into physical instantiations of robots, wearables, cars. Uh, so that's the physical AI is becoming a big part of the future conversation with, you know, that's happening. Uh, the new protocols that are emerging, you know, the model MCP, the model context protocol, creating a new protocol for AI agents to work together, you know, you know, creating new standards or working together across the, you know, uh, agent ecosystem, if you will. And I think all of this is creating a flywheel effect for organizations to harness, you know, as we think about the vision for a GI or the super intelligence, you know, it's creating the flywheel effect. And it all depends upon the organizations of how we can, you know, get this flywheel effect and embed it into the organizational strategy. I think that's the big opportunity. Having said that, I think the big challenge, you know, and this is something probably is there in many of your minds, right, uh, scaling this ai, you know, having said the opportunity is hard. It's really hard. the recent research that Accenture did around Front Runners Guide for scaling AI shows that only 8% of the organizations are able to scale AI and embed it part of their business strategy. And, you know, as as we looked at, you know, our 2000 plus gene AI programs that we did, that we have done with clients around the world, there were three, four things that came out as challenges. Number one, as foundation. It can be, it was about data being ready. Is the data in your organization ready so that you can actually scale the ai? That's, it's as fundamental as it can be. Big, big, you know, question marks around do we have clarity around the big bets that we are driving with ai? Do we have alignment within the organization around what are those big bets that are business led, you know, as part of our AI agenda? The third big, you know, question mark is, is our talent. Is the culture of the organization. Ready is the mind, is the do the, do the people have the right mindset for us to adopt ai? You know, it's not just the technology. So do we have that and do we have the right leadership and sponsorship for enabling that right culture, sustain that culture and si sustain that mindset, walk the talk. So that's what are going to talk today. You know, I wanted to give that context. That's what we're gonna talk today and as, uh, was, you know, mentioned in the introduction. I have the privilege of working with Dale over the last couple years and, you know, I will invite Dale to come and talk about her journey, you know, and, and the, the newspaper clipping speaks a little bit of that. So, over to you, Dale.

Telstra's AI Strategy and Collaboration

Reimagining Business Value with AI

Speaker 4 8:31

Thank you. So this newspaper clipping came from the Australian Financial Review, so think Wall Street Journal. It's Australian's kind of corporate financial paper. And this article came out in January just as we were announcing the joint venture with Accenture and the A FR tells us that it was their most, Clicked article in, in the year for, and, and drove more subscribers to them. So they love us at the moment. They like talking to us all the time. Um, but it is representative of the big bet that Telstra made, that Telstra is making on ai. And so we had been working on AI for years. We have hundreds of AI models in production already, but really we were looking at what our 2030 strategy was, the ambitions that we had in it, how Telstra as a business was going to rely on ai, uh, not to become an AI company, but to, but, but the business that we are in and what its reliance on AI was going to be. And we knew that we couldn't do that ourselves. That we needed to, we needed to work differently and think differently about how we were doing it. And that's where the announcement on this joint venture with Accenture is. But who are Telstra? So, I mean, uh, in Australia, everyone knows Telstra. There are about 27 million people in Australia, and you can see on here that there are 20, somewhere, 24 million different, retail mobile services that people have out there. So we have a very large, we capture a very large part of the, of the market in Australia. Our, our main purpose as a business though, is connectivity. We run the network, we run the telco network in Australia, that we have competitors, we do have competitors. but our, our role is really connectivity Australia. You can see that little drawing of Australia up there and where those dots are, they are, I think they are overestimating where people live in Australia. Most people in Australia, there's 27 million people live down the East coast and most of them are in, are in Melbourne and Sydney. beyond that though, we do have to connect to the whole of Australia. And so that connectivity in very remote regions is, is what we are all about. And the way the network works, is. Is that, that reliance and reliability of the network is really, really important. We also run, um, emergency services. So triple zero in Australia is 9 1 1 9 1 1. Is that right? Yeah. 9 1 1 here. and so we run that. So it is, you know, from a healthcare point of view, emergency services point of view, that reliability on connectivity on the network is huge. Telstra also has, uh, the Telstra Foundation, a philanthropy part of the organization, which is really aimed at digital inclusion for everybody. Digital inclusion is really hard if you can't connect to the, if you can't connect. And so we, we play a really big role in being able to connect the whole of Australia so that everybody can thrive as, as the world changes around us. So that's a bit about Telstra, who we're, But, so let me talk a bit more about, uh, data and AI at Telstra. So, a, a few years ago, we, so we, we had data, we had AI in our company strategy. That was a three year strategy that went from 2022 to 25. And how AI was represented in that strategy said we were going to use AI to improve a hundred percent of our key business processes. So if you stop and unpick that a little bit, it meant we had a set of business processes and we could see that AI could be used to improve them, make them more efficient, make them run faster. but it didn't think about how AI might reinvent our business. And so the way we went about doing that was literally that we shoved AI into all, all of our existing business processes. They went faster. They, and all of the things that we wanted were happening. But when we looked at our strategy going forward, we could see how much connectivity, Telstra, our, our, the big things that underpin our strategy, our network resilience, our customer experience, and our employee, our employee experience as well. and how AI was gonna play a role in those, wasn't just making our existing processes go faster. It was thinking about in the age of ai, how did we need to, how could we reinvent the company? And so that's what we set out to do. All boats rising on the same tide. I actually heard a couple of people talking about that yesterday, when you think about society, but in Tide, a company like Telstra, I think for, for those of you that have worked in large organizations, there are silos in those organizations. The people that you know are, are retail customer facing versus the people that run your core product like our network, or, you know, the different divisions. They, they, they silos exist and everyone talks about we will collaborate and we'll break down the silos and no one ever has. And so, uh, this, all boats rise on the same tide is super important when it comes to AI because, uh, if you are reinventing the company, you can't reinvent parts of it. You need to think of the whole, and so when we formed the Telstra's data and AI strategy, the very first key to that is that there is no data and AI strategy. It's a company business strategy. So what is the. Our business strategy for Telstra and what role is AI going to play in that? And that really went a long way to it being not so much, I mean, it is still a, a tech led, AI led reinvention, but it is done from the purpose of not me and a NA pushing AI onto everyone, but really the whole organization reimagining what they might do differently with ai. But it didn't start like that. I mean, we are, uh, we'll talk a bit more about what some of the big transformation use cases and how we've brought the whole company together on those. But it didn't start like that. We were absolutely in silos. People were interested in ai. I was really worried we were going to scale AI use case by use case by use case, which was just gonna create more fragmentation for us to deal with and more risk. And so it was in September, 2023 where we had all of our enterprise leaders together. So the top 200 people in the organization, and we were talking about AI and Australia also has this. Like more so than anywhere else, I've seen this real concern about AI replacing people's roles. It's in all of the media, it's in, like I haven't seen the same level here at all, but it's really, it's really big in Australia. And so we were sitting with this with our enterprise leaders group and everyone was getting a little bit excited about ai, but really nervous about it. And we really wanted to get to this. All boats rise on the same tide. AI being the company strategy. and so I said to that group of people, AI is gonna replace roles. And the roles it's gonna replace are the enterprise leaders who aren't thinking about how ai AI is gonna come into their business. And that's all of you. And it's also really good when you need to, for people that attend a lot of meetings, who read and write a lot of documents, and that's also all of you. And so it's coming after your roles first, but so don't think it's coming after the junior people's roles. It's coming after your roles. And so you have to start using it. and the way to start doing that is to play with it. use it yourself. Play with it yourself. Work out. Use it in like just for play or use it for your own, your own work purposes. And once you've got your head around it, then you can think about how it's gonna help your teams and how it's part of your strategy and you're not relying on other people to tell you how to do things. So play with it first. And I promised someone last night that I would, I would, introduce into this conversation. Um.'cause I am Australian. but so one of the key premises of b bluey is how you learn through play. And so, all of the Telstra leaders then are learning through play because they're experimenting with it themselves. And I've heard our group executives talk about how they've made a, you know, what would be my movie headline for this, my work this week, and stuff like that. So they are playing with it to experiment with it, to work out how to bring it into their teams and into their strategy. And so that's what's really happened. We've really, I think one of our key premises was that we wanted the AI strategy to be a business strategy. and we've kind of ticked that one off. Actually. We haven't ticked any of the other ones off yet in terms of our goals. But that, that first one we have. The next part was when we looked at the ambition in our, in Telstra's strategy and our, our now 2030 strategy. Then there is an extraordinary amount of ambition about what, what, what we wanna achieve with ai. And it's not just being, you know, more productivity, more efficiency, it's actually how are we gonna reinvent the network? So an autonom, an autonomous self-healing network is kind of a goal, right? So if our, our network today has over a billion data points coming into every day, coming in every day about how the network is performing. Humans can't, humans can't look at that much data. they're looking at where the red flashing lights are, where it's broken, where, where things aren't working. but they're not looking at what are the patterns that are indicating a, an a, a fault is gonna happen. They're not able to look at how did we fix that fault before and, and how might we fix, can we fix it remotely now? And so that autonomous self-healing network to be able to ensure that connectivity for the whole of Australia is one of the big goals there. But that same kind of goal is when we're thinking about what customer experience, what customer service is gonna look like in the future, what our employee experience is gonna look like in the future. All of that reinvention and all of the ambition that we have in that, is like we, we realized that we just could not do that by ourselves. AI is, you know, being a, a leading company, the size of Telstra in Australia with the ambitions that we had on AI meant that we needed, you know, our. We, it was a global game for us to be able to really, really, really bring that AI expertise in to help us drive that forward. And what we were sitting on, and we'll talk about that a bit in a minute, is that we had, we had, we, we have a lot of, you know, legacy IT systems, legacy data systems, all of these kind of things that we needed to accelerate forward as well. When we looked at our roadmaps to say, what, where do we need to be to achieve that AI ambition we had, it was gonna take us five years to get there with the corporate settings of funding and resourcing and everything else that we had. And five years is too long. Five years in the world of AI means that we'll always be five years behind. Um, and we'll create more risk and we'll, we'll, we'll lose opportunities that we have. So we needed to go faster. And that's where we started talking to Accenture. So we went with, we, we wanna, like, we started with it. We didn't start with the premise of creating a joint venture. We started with the premise of we wanted to accelerate, we wanted to get five year roadmap into two. And, and, and deliver extraordinary amounts of value for us as an organization in our applied use of AI as well. And so as we worked through that, we tried a few different things. Um, we did some like individual kind of statement of work type engagements to try to accelerate parts of what we were looking at. And that worked a little bit. It didn't, it didn't, it wasn't gonna get us from five years into two though. And so then we, we, we looked at, at different ways of doing that. That's where we ended up in, in the joint venture. And so that joint venture now is, we have. About 1200 people working in the data and AI team at Telstra. Um, like specifically the data and AI professionals, they were about 50% were Telstra employees and 50% were, resources from 40 or so different vendors. so we consolidated all of those into just one with Accenture. Joined those two teams together and, and really looked at how do we accelerate then how do we bring in global expertise, how do we open up, it's, we are a really long way away. It was 27 hours just to get to Chicago. So nearly 30 hours to get to here. We are a really long way away from everybody else. Um, and so that. That access to global talent, coming to work on our problems is, is actually quite difficult. and so being able to like setting up this joint venture with Accenture was quite a different model for Accenture as well. Um, meant that we have someone like a NA their chief, you know, ran data and AI capabilities for Accenture in Americas and is a global chief responsible AI officer, and then an NAB's leadership team that's come in, have come from all over the world as well, and bringing this expertise as well as a whole team of about 600 people that have joined us, bringing this expertise that we just did not have access to before. So that as well as the, the tools and a whole lot of other accelerators that an unable talk about shortly is really what that, what that accelerating with allies has been all about. The other thing is, is the way that we partner and collaborate with other tech vendors. So we, we work with Microsoft, AWS, Salesforce, SAP, Workday, all of these, all of these, uh, all of these different organizations, Ericsson and others in the network space. the way we partner in Australia. And even with, with universities, I mean, I've heard conversations here today with, um, Steve this morning talking about how you, Notre Dame collaborates with industry. It's quite a different conversation in Australia. We don't have that deep collaboration. and when we look across, across our technology vendors that we have, we have really close strategic partner relationships, but they're very one-on-one. So in just our customer relationship, part of our business, we, we use Salesforce, a lot of our customer data, then that sits on AWS and then we have different vendors in there working in that space as well. But we don't bring them all together to say, this is our vision for reinventing customer service. Bring them all together to say, how might we work together on resolve, on, on reinventing this. We go and have a conversation with Salesforce and get their ideas. We go and have a conversation with AWS and get their ideas and then we try to solve all of that. Um. That is not how it works in the us. So every time we've been, I've been over here, I hear so much about the, the collaboration and innovation and partnerships that happen across, across between academia and industry and the different industry groups come like, groups coming together to create these different solutions. And I think an, you'll talk about trusted agent huddle and different things where these partnerships are really deep and they're really driven to drive industry forward. We just, we don't have those kind of relationships in Australia. And so that was another thing that we really wanted outta this joint venture. How do we start to collaborate and partner with the advanced engineering groups of the universities that Accenture has relationships with and the other vendors like Salesforce, AWS, Microsoft. How do we get that collaboration working differently for us in Australia as well? So. Global access to talent, global, uh, access to a lot of the AI acceleration tools that Accenture have. And we have set up a hub in Mountain View that is like specifically drawing advanced engineers capabilities from all those different organizations to us to work on our problems. And I keep hearing, some of Telstra's customers in Australia and some of our peers wanna go and visit this our Mountain View hub. I'm like, it's not an executive briefing center, it's a team that is just working. There's like a poster on the wall that says Telstra, but otherwise there's nothing to see there apart from people that are working on our problems. So accelerating with allies was a really, was a really big part of changing the game about how we work with, to, to really, really go after this AI ambition that we have. after spoken about all that, I'll pause on that for a moment, but I have covered all of that. and so when we then set out to say what are we working on? How might we break this down? the, these are the things that were really the, the building blocks that we went after the, I said that we'd, we'd ticked the box on one that we needed to have a whole of business strategy. So we used to have six building blocks. Now we've only got five.'cause we've ticked the box on AI strategy being a whole of business strategy. But the other areas are that we really need to go after unlocking value. We're not just building tech and we're not just building AI for experimental kind of AI sake. It really needs to be going towards business value. So we've kind of gone from that. AI needs to be a business strategy into that. Business strategy is about unlocking business value. And so, we'll deep dive on each one of these as we go forward. Modernizing data and insights was really looking at the foundations that we're sitting on. So what do we, and I said before, like the, the Frontier Front Runner Company's really got, they, they've got their data in order. They've got, you know, all of these different things in order. So that, that modernizing our data and insights is a big part as well. and then step change in our AI capabilities, going from that scaling use case by use case by use case to how are we building foundations. That means that we can plug, use cases, plug agents, plug all of these different things in really easily. Um, and we'll go through a bunch of examples on that as well. And then responsible ai. So Telstra in Australia has long been front runners and leaders in responsible ai. We were, part of the original pilot program with the Australian government to put the Australian AI ethics principles in place, which there's eight of them. They are similar, but. I mean, they're pretty much the same actually as the, AI value, uh, stuff that was up in the health, presentation last night. But those eight principles of our responsible ai, we were early adopters and co-developers of those with the Australian government. we also work with the GSMA, so the global mobile industry to help. We've co-authored their AI for good playbooks and they're the global mobile industry's way of thinking about, uh, responsible ai. And from all of that work, we've been invited to be, have a seat on the UNESCO Business Council on AI ethics as well. So all of that is great in theory in terms of how we're thinking about responsible AI in Australia, because Telstra is such a large organization and we have contact with nearly every person in Australia because of our network and our reach. Then what we do with AI and the decisions we make about responsible ai, whether it. Impacts you as a customer or not matters to how the rest of the Australia operates when it comes to ai. So if we take a shortcut or if we say that doesn't matter, then it gives permission for everyone else to do the same thing. If we hold ourselves above that, like the, but we, we run ahead of the regulations.'cause the regulations I think are the same here. The regulations aren't moving as fast as, as how as, as AI is changing then the way we hold ourselves accountable for responsible ai, regardless of, not regardless, but ahead of the regulations matters for the rest of Australia. So that kind of weighs heavily on our shoulders, that responsibility that we hold. but all of that theory in terms of a AI principles and frameworks and stuff like that, only matter if you can operationalize it into every single employee for every single customer, for every single person that's building AI in Telstra. Um, so that opera operationalization of responsible AI is a big part of what we wanted to accelerate as well. And then people, we'll come to that later as well, because AI isn't all about tech. You only get the business value. You only get the change when people adopt it and people use it. And so a really large part of what we're doing isn't about tech at all, it's about people. And so we're gonna double click on all of these. I'm gonna start with value because we have to work really hard on re-imagining how we calculate value and what business value from AI means, and then will, um, jump more into some of the acceleration work we're doing around data and AI and responsible AI as well. So reimagining re-imaging, re-imagining value that is supposed to say, the way, the way most organizations think about value and think about what they choose to invest in is a, uh, is a kind of, oh, uh, I call it old school. It's that old's called,'cause it's current for most people. is a calculation based business case. This is how much it's gonna cost. This is what the benefits are gonna be. That means that this is a return on investment. It's a hurdle rate of, I dunno, 11% means you get a tick and you get the investment money. AI just doesn't work like that sometimes. It is quite experimental at the start. You don't quite know what the, what the outcome's gonna be. You've got this data, you've got an idea, how's that gonna work? So we, we needed to take how most people work when it comes to innovation. When you've got an innovation idea, how do you, how do you, how do you work with that to get to the, you know, the. Pivot, pause, persist type conversations, lots of stage gates around it so we can get to that value calculation at some point. So really, working, really working out how we have investment in AI from an innovation and experimentation point of view, while having strong guardrails around it was the first part. The second part though, is there's some parts in AI where you just don't know what that benefit's gonna be, but you can see there's gonna be a benefit, but you really have to work it. So Telstra, are the larger, we've, we've had it for 18 months I think. we are the largest, corporate in Australia in terms of Microsoft co-pilot licenses, Microsoft talks. I dunno whether there's anyone in from Microsoft in the room, but Microsoft talk a really big game about, you know, the benefits that you can get an organization can get from Microsoft Co-pilot. It's really hard to bank those benefits though. And so our CFO wants to, you know, like we need to that value and that that benefit calculation needs to be bankable to our chief financial officer and Microsoft copilot licenses aren't cheap. and it's really hard to get to some of the outcomes that they talk to, but we had to make a decision to say, to like a conviction based decision rather than a calculation based decision in that investment to say we are going to, we're gonna go all in 21,000. We've got 28,000 Microsoft co-pilot licenses. Now that is actually more, nearly more employees that we have, but I'm not quite sure how we did that deal. But the, the, the like making that. Making that decision without really knowing how we would get the benefit or how we would bank that benefit, means that we had to work really hard with the, you know, the financial decision makers in our organization about how we make conviction based bets on ai. But we don't, once we've made that conviction based bet, we still need to go and then work out how are we achieving value on that. So if you think about the way, I don't know, your investment committees work or the way your financial teams work, they, they, they kind of wanna know, they wanna know that they can bank the benefit before they'll give you the money in the first place. And so we really had to work to reimagine. And reinvent with our finance teams, how we would do this, you know, make, make decisions on what we would invest in when it came to ai, and then how we would measure it. So we've also worked really hard on a, on a value assessment framework so that every a piece of AI work that we are doing gets assessed in terms of what, what we think the value is gonna be. And it's not just like a technology kind of project where you, you build the thing and you put it out there and the benefits come in. The benefits don't come always, sometimes, not always from ai just because it exists. It comes from how people are using it. And so that then leads to a very different conversation with our business owners. So AI being that business strategy when we talk about there is however many millions of hundreds of millions of dollars of benefit in AI that we are building at Telstra this year. that benefit comes from the decisions that the business units make on what roles their employees will be doing, how their employees are using it, and, and, and really pushing the adoption. And then, you know, if they have to make decisions that mean they move people out of that section into another section, whatever. But it comes to a business decision, not just because we've built the tech. And so, um, this, this re-imagining the way we work with value, the way we measure it, how we make decisions around it, and how we ensure we get the benefit in the end, has meant a very different way of working when it comes to AI than most of the rest of. we've also worked with our business on, like what are the key transformational a like we have, I think we have over a nearly a hundred, individual kind of AI use cases that are in, in development this year, building on top of all of the ones we've had in the past. But we really wanted to, but, but a lot of those kind of still sit in that space of either they're quite experimental or they're improving the way things are done today. They're not really thinking about how are we reinventing our, our, our organization. And so these, we sat down with our group executives and our CEO, Anna and I joined that, that, that, that offsite, a whole day on AI where we really, we really adjudicated what were the, what, what were the, transformational initiatives we were gonna go after. And so how, um, agen AI will play a role in customer care. So how do we reinvent customer care, fixed inventory? If you think about our network sitting all around Australia, just the parts, uh, that, that fixed in inventory one is kind of the first big step towards autonomous network. We have to sort, we've got an inventory, an inventory management problem that we really need to solve. And so we're applying AI to that. AI assisted migrations, we are still moving off Legacy, um, Siebel kind of based and even older, customer relationship management systems and the way that you migrate customers in Australia onto a new system is very disruptive for the customer. And so it's not just about do we have complete data sets? It's all sorts of regulations come in play. And if you think about you're a customer and. Mobile phone provider is changing their IT system, which means you need to get a new phone number and you, your billing gets disrupted and all of these different things, like we don't want that for our customers. It's very messy. And so we're looking at how we use AI to make that better because it hasn't been a great experience and it's taken us far too long to be able to do it. B2B sales reinvention. So how are we helping all of our sales agents do their jobs differently with ai? Um, and then digital twin for the entire business. I feel like digital twin conversations have been around for a while, but a big part of our 2030 strategy is, uh, re-imagining network as a product. And so most telcos, the networks and network and the danger for telcos at the moment is that people think of as a utility, something. Network speeds in Australia are much worse than here, actually. But if you think about, you're a small business owner, you're at a festival, you wanna guarantee that you can take payments from someone and you don't have connectivity that day, then you would pay more to have guaranteed connectivity or different, depending what business you're in, you would pay different amounts to ensure different. Different speeds of your network and stuff like that. So we, so we are, we are reimagining the network as a product as part of our strategy. We have been working for years though, to simplify our business. I think five years ago we had like 1800 different products on the market. We now have 20. And so we have been re radically simplifying our business. The danger with network as a product is we blow up that complexity again with every little, you know, tweak we make to the network and what we sell to people. And so what we wanna do with Digital Twin is, is really specific to that. We, we want to make our business with network as a product. We want to become more sophisticated, not more complex. And so we wanna use Digital Twin if we make this tweak over here, if we change pricing here, what does that mean over here in product economics and all of those kind of things. So Digital Twin is really big for us. Fraud, ai, customer iq. And then, fixed network are, three other big use cases that are really specific to different parts of our business as well. So we're really, the organization has really come behind. These are areas where we wanna, where we are first starting to reinvent what our business is, um, with the use of AI and with the help of the joint venture we have with Accenture. But all of that only works if we start with the foundations. You cannot build any of this on sand. We had very complex technology environments. We're a 200 year old company. Oh, uh, ex federal government company, that privatized 20 years ago. and so very complex technology systems, which meant we had very complex data. When I started in the role, three or four years ago, we had over 80 different data warehouses, data platforms. we have. People trying to get their hands on the data that they need, dumping it into data Mars, like there's data was, data was very messy. We're down to 30 platforms. Now we've gotta get to three, data. But in terms of, we, we wanna, we need to protect our data more. There's been some really big, messy data breaches, privacy breaches in Australia. We, we wanna, as we simplify our, our landscape, we want to protect our data more, but at the same time make it more accessible to the right people to be able to do, you know, the right things with analytics and ai. So there's a whole lot of work that we need to do in the data space, and, and, and, and beyond actually. But, ANAB, I'm gonna invite you up to talk a bit about how we're working with Accenture to really accelerate this space.

AI in Retail Store Management

Speaker 3 35:00

Sure. No, I think, I think this is great. You know, the conversation that Dale just had was really around the business strategy and how we are anchored on the business value. But, you know, when Dale and I, and you know, we were actually working with all the, you know, group CEOs and the chief execs at Telstra, what we realized was that if you don't build these foundations. We are not going to win the trust. You know, and that's where the foundations are extremely important, so that we can build the confidence, you know, both within the organization, with our employees, and for ourselves, that we can deliver the promise that, that, that we have made. Right? And, and as I highlighted that data, you know, is, is the most critical asset. And how do we do that? Like, as Dale mentioned, we had. Plus data platforms, legacy data platforms, you know, the likes of Teradata and others. And we have brought it down to 30. And the, the plan is over the next 12 months, we'll be actually going down from 30 to 15 and then finally to three. So that's kind of the, the, the accelerated roadmap. And as part of that, what we are doing is we are in the spaghetti that you see on the left hand side, right? That's the spaghetti of the legacy platforms. And I'm sure everybody can relate to that. It's very hard to actually ma mass the data from one system to the other system, get a unified report, you know, that you can trust. So we are trying to actually simplify that, that spaghetti mass, you know, migrate that into cloud, you know, and, and Microsoft is our, our cloud partner of choice. We're, you know, working through that, but then modernize that whole platform, you know, and we're, you know, bringing in new innovations around the whole lake house architecture that helps to bring the best of breed from the, you know, the data warehouse and the data lake concepts, bringing it together so that we can go multimodal when it comes to data processing. We can do data sharing, like for us, sharing customer data. Along with the network data is extremely critical so that we can actually be very proactive, understanding the network, you know, fluctuations and anomalies to understand what needs to happen with respect to the customer experience, you know, in a very, very proactive basis. So sharing of the data is extremely critical, is really important for us to have a unified governance of that data. So the modern architecture is going to facilitate the unified governance across the entire data life cycle. And then what we are doing on the back of it is creating data products. Think of it as reusable data assets. Hundreds of them that are curated, that are governed, that are trusted, and we are deploying it on a marketplace, an Amazon like marketplace for the entire company so that you know, business stakeholders, whether you are a finance professional or whether you're an HR professional, you're a sales professional, you have got now those reusable data assets that you can trust. You can use those data, data products to make business decisions. And then what we are doing on top of it is to create the underlying context, you know, with the metadata, the data lineage, the underlying data quality, so that you can really enhance the trust as well as you can protect. You know that, that particular data in the, you know, with the, with the respect that you need to provide, with respect to the regulations and the compliance to the laws of the land as well. So this is a massive, you know, effort. And as Dale mentioned, you know, this would've taken five years and we are bringing in number of accelerators and assets from our Silicon Valley hub to make that happen over the next couple years. So you're already in the journey. We are seeing the results, as I said, 80 to 30, 30 to 15, 15 to three. You know, it's moving at a pretty fast pace. What it enables us to do is now on the back of this data foundation, is enabling us to create the insights and re-imagining how insights are served to the business. You know, this is a humongous uh, effort. You know, Telstra has over, you know, 45,000 plus reports that gets created on an annual basis, and people look at those reports, you know, they have some skepticism around those reports. It's very hard. So we are trying to reimagine saying that you don't need those thousands of, of reports and dashboards to make a business decisions. How can we create an AI over BI capability that reimagines how you actually not become reactive to looking at a dashboard? Can we now be more proactive around those business insights? By creating those insights, the alerts, the automated summaries start to bring gene ai, you know, to, you know, and semantic search capabilities. So then you can have a deeper dialogue with the data. You know, and, and that is very per, you know, personified based on whether you're a finance professional or a sales professional. It, it's persona specific. And then how can you start to, you know, use AI to augment the dashboards that is, you know, much more action oriented, right? How can you do that? And we are, we are starting to see significant impact in the way we measure performance, you know, across the business. One of the areas, for example, is the retail stores. You know, the Telstra has its retail stores. Customers like us will walk into that and a store manager who is managing the store has to manage the performance of the store. You know, typically what they do is they look at 40 plus different reports beginning of the week to do the planning, understand how the performance was last week, and based on that, make specific actions through the days of the week. That's a humongous process in annually spending about 20,000 plus hours doing that. How can we change that paradigm using AI over bi? And the reception that we are getting from the, the store managers has been phenomenal. So I want to play a quick video about how we are reimagining that on the back of data Foundation and the, and the business insights. So what you'll see here in this, in this in a quick demonstration is a store manager, you know, looking at understanding the performance of the store, having a dialogue with the data to understand the root causes, taking specific actions, including coaching, you know, of the store agents and so on, so forth.

Speaker 5 40:22

When Barbara logs in, she's instantly greeted with performance results for her store on the left. The business summary provides a snapshot of overall performance along with key highlights. She can copy and share directly with her team via Microsoft teams. Just below the performance highlights, show her RBS results for the week. Barbara notices a drop in her store ratings from four to two. The decline is mainly driven by low scores in reputation and product scoring, just one and two respectively. Concerned about the sharp drop in reputation. Barbara dives into performance insights, which is powered by the diagnostic capabilities of AI driven insight. She discovers that reputation fell due to three detractors and one passive response in recent surveys. To understand more open, a quick access chat assistant built using gen, AI and large language models to enable Barbara to have a dialogue with her reputation data, it reveals that Ken, Chelsea, and Midge were the agents linked to those survey results. Barbara wants to dig deeper. What exactly did customers say? She types in a question. What were the key themes from the verbatims? The assistant responds with a summary. Customers were confused by poor explanations, unclear recharge instructions, and long wait times. Now she's curious, has this been a recurring issue? She asks for a breakdown of survey distribution across her agents and a deeper look into NPS trends with one click she's taken to the NPS analysis board, which auto generates insights based on frequently asked questions, also known as prompts. These analysis boards can be customized to any repetitive analysis done by an end user. The results are telling Ryan is performing well, but Chelsea, Ken, and Midge show inconsistent results with Chelsea, consistently underperforming, and Ken and Midge showing a sharp decline. That's something to note from a coaching perspective. Additionally, mobiles as a product category has been underperforming with conversations consistently receiving low NPS scores. Later, Barbara returns to the home screen and notices a separate proactive alert. Her unicorn offer usage is currently 30% below regional average. This presents a clear opportunity for improvement. This brings together descriptive performance data with proactive alerting to give Barbara an actionable insight. The data indicates it's directly tied to sales with both impact and criticality rated high, making it a priority area for immediate action. In the driver analysis section, Barbara sees that unicorn offer usage is 30% below the regional benchmark, and it appears to be one of the most significant contributing factors to close the gap. The recommendation is to use the offer lease three times a day. This recommendation is driven by machine learning capabilities, looking at historical data, and is able to recommend a clear, tangible goal. She can cascade to her team to understand how to act on this. She navigates to linked actions where she finds recommended steps to address the alert. These include coaching opportunities and learning resources available across other stores in her region. She clicks view details under coaching, which identifies the specific agents who would benefit from support, outlines the reasons they were flagged, and connects it directly to the alert and offer in question. That concludes our demonstration on AI for bi.

Speaker 3 44:07

So, you know, that kind of gives you a sense as to bring it to life, you know, as to what does a real solution look like. But obviously, you know, this is still, you know, at the data and the insights. And now what we are doing is we are scaling up the AI foundation on top of this. Right? And for us, you know, it was really important to think through what are the architecture principles as we think of scale, scaling the AI. And as Dale mentioned earlier, that, you know, we want to get out of this paradigm of use case after use case, starting to build that, how do you actually start to think about reusable modular components, you know, across your AI foundation that can scale? That was extremely important for us because we are building it for the whole company. Uh, speed was extremely important. You know, how do you, how do you, you know, build it in such a way that you can, you know, accelerate the speed at which you're deploying the solutions, AI solutions. And then resilience is extremely important so that the AI systems that you're building are resilient so that it can give you the confidence and the trust. And as part of that, you know, you see the six. Six, uh, images here, these kind of forms. Our key, um, building blocks of the, of the AI foundation. You know, we are building up the entire gene AI and the model and the foundation model capabilities including, you know, model selection, model switching capabilities, model orchestration, LLM orchestration capabilities. So I think that's, you know, part of our GI foundation. We're building a whole, you know, company-wide agent, TKI platform that also enables us to build our own agents, but also leverage the agents that are coming from enterprise systems, whether it's Salesforce or ServiceNow or SAP. And, you know, with, uh, you. What we are, what we are bringing together is like a trusted agent huddle that helps us to orchestrate these agents. You know, there are super user agents, there are utility agents. Bringing all of those together to reimagine a specific business process. Uh, we are activating a whole decision intelligence capability as well, you know, that has simulation capabilities and, you know, enabling us to take the, you know, next best actions. Uh, we are establishing the pipeline, you know, for. Model deployment model observability with, you know, ML ops to make sure we can govern those models across the lifecycle, you know, uh, of, of the model. Um, something called as the control plane, we are establishing, which is to make sure that we have guardrails established to, uh, take care of the model performance, you know, to make sure that we are working around the ethical guardrails that are important for the company. And then consent and preference, you know, with respect to customer privacy is extremely critical as we think about the personalization at scale. How do you, how do you, you know, embed those frameworks as part of your ai, uh, AI foundation? So those are, you know, couple of highlights. And what it is enabling us to do is, you know, build out these AI solutions that helps our employees, you know, helps our frontline, people. And one of the, uh, successes we had is called Ask Telstra. And Ask Telstra is a AI solution that works across 8,000 of the frontline, you know, uh, individuals at Telstra who are using that to answer customer queries. And, and, and using all the JEI capabilities and the knowledge base to, to address that. So I wanted to give a quick, uh, preview. I'm just wondering if the time is enough. Nih, do you have, uh, you know, another five plus minutes? Yeah. So I'll just play a quick video just to give you a sense of, uh, what the Telstra solution looks like. So this is just an example of one of the solutions, you know, that frontline agents are using. It's saving, you know, almost a minute for each of them, you know, for the call. And, you know, we take millions of calls and million, and this is such a huge enabler for, for them to boost their own productivity. you know, and there are number of those such solutions that are in play right now at Telstra and Dale, maybe you wanna provide a comment, you know, around the solutions that are even before the AI actually.

AI Adoption and Employee Training

Speaker 4 50:08

Yeah, so we have had AI solutions in place at Telstra a really long time. I mentioned before that we were embedding AI into a lot of our, a lot of what we were already doing. But, I, I think we, that the, that reimagining those instead of going, that use case by use case has been really important for us. So, how we've used us, Telstra, for example, and how we've built on a bunch of these that were there in the first place. So, has meant like we're really trying to move from those. Use case by use case to scaling and building from foundations. Um, but we have, we have AI sitting across our network, career planning, everything that, like, lot, lots of customer facing things as well. So, it's everywhere at Telstra and our job is really to harness it and be able to take it to the next level.

Speaker 3 50:47

Yeah. So, you know, I think, I'm just looking at the watch here. I mean, we covered a lot of the technical foundations that are needed actually. Right. But this is not only about the technical foundation, it's a lot about the people and the mindset. So maybe Neil, you wanna just cover about the efforts we're doing around the people and the mindset and, and, uh, and the message on that, and then we can close. Yeah.

Speaker 4 51:05

So I think, I spoke earlier about, you know, the value from AI really coming from how people are adopted. Those two examples that a NAB went through. Store managers, store managers at the moment show up to work with about 40 different dashboards that they have to look at to get that same information that was in that short video. It's taking them up, you know, the really good ones. It can take them half an hour or an hour and a half every day. And so we are not reducing the number of stores we have. It's, you know, AI is not threatening store manager roles, but they really needed to be able to go forward to move faster. So that example Anab gave about the store manager, AI for BI that they have is taking that hour and a half down to less than 10 minutes at the start of their day. And so if you're thinking about how you capture the mindset, how do you capture someone to want to adopt the AI to really wanna build it, then going after their biggest pain points and helping them be able to do their job better is the way that we're really working to capture the minds, not the machines. With the US Telstra example in our company engagement survey. The, the year that we, we built that as Telstra. I mean, you can see it just looks like a chat. GPT, that's really kind of like, you know, tuned to our pri our pricing policy procedure type manuals for our frontline agents. It's, it's not super complex in what sits behind it, but, it was that access to knowledge, quick access to knowledge for our contact centers and our frontline staff was the biggest pain point that they called out in the engagement survey that year. So being able to. Point, the, the, you know, our big kind of bets on AI to pain points, like the, the biggest pain points that our store managers and our frontline agents have, gives us a much better chance of them being able to them adopting it and changing their way of working, learning new skills to be able to, you know, write prompts rather than, kind of just knowing where that data was kept before, um, has been really important for us. But it's, it's that, that's kind of the, that trick I think. Like, how do you, how do you for, for like a large organization and a large kind of. Space of people that you need to adopt it, how do you point it at their pain points? So you're helping them to do their job better. Alongside that, we've built an academy, so we know AI will disrupt, disrupt roles in our organization. but we have also invested in an a data and AI academy that is persona based, curated learning pathways through lots of online content that every single employee in Telstra has access to. And one of my proudest moments, I think is I heard third hand of, from someone at a dinner party that, um, there was an employee from Telstra there and they were taught, I have no idea who the person was actually, but they were talking about, you know, what do you think about ai? Is it, is AI gonna come and replace roles? And they're like, absolutely. My role, my role could completely be replaced by ai. But I have access to this learning pathway and I can go and I, like, I, it's in my hands as to whether I, I go and learn and adapt to, to think about what my role is gonna be in the future. So I was super proud of that because that was a little academy that we built before we partnered up with Accenture. We kind of built it based on some data camp licenses and about three different people. And now we have access to all of Accenture's learning and development and their, and their partnership with Udacity. So we can really blow that up and make that, that, that super powerful for all of our employees. So that minds not machines is, is really that pathway to getting to value and thinking about how you're gonna capture the mindsets of people. It's a little bit about our academy. and then the last part is that it's all changing so fast. You know, we all, I mean, we're all, we're all here at a, at an AI conference, and I spoke to some different people yesterday that were like, I'm here to learn. I wanna know more about what's going on. It's changing so fast. Organizations like us, you tend to work in annual investment cycles, so you place your bets at the start of the year and you just have to go deliver that. And then. You know, some something, something happens and you have to be able to change it really quickly. So, being in large organizations that do tend to move in annual cycles, being able to adjust what you're doing as the world changes around you is kind of dependent on having all of those things in place that, you know, that if we make this change, it's gonna lead to value. We've got the foundations in place first. We can pivot really quickly. Being able to be, you know, a big old organization that can pivot quite quickly is really, is really important when we're working in a space that changes as fast as, as fast as this one is. So, um, ANAB spoke at the start about the frontier AI Frontiers, report that Accenture have put out. you can find that on the, you can find that on the internet anywhere. It's, it's a really interesting read, but those imperatives on the, on the left hand side there of. On the left hand side, um, really like, they come from that report. And there are a lot of the things that we are really thinking of as well as to how we bring those in at Telstra and what we're trying to do. And then our learnings as we've gone along. You know, AI being a whole of business strategy, not a tech strategy. that we really have to capture the mindsets, not just capture the people to be able to get to that value. We have to think about the foundations. We, we need to think differently about how we partner with different organizations and we're trying to do that. And then being ready to adjust and, and pivot really quickly when as, as we work in such a, in a space that's so full of fast-paced change. So that's, All we have for you today, you can contact us anytime.

Q&A and Closing Remarks

1 56:15

Thank you so much. Uh, Dale Anona. Ben and I couldn't agree more with you. I think tying it with the business objectives is absolutely essential because otherwise, you know, it remains a curiosity and efficiency approach and we need to attach it to a business outcomes approach. That there's not only cost savings, perhaps in automating our processes, but also business revenue that can come in because we realize the cost savings within the first one year, perhaps what happens in the second day. Right? Uh, and so I think, so I really, really appreciate it and you know, the pillars of, and I loved your point about. Updating minds and not just machines. And, and I was looking at your dashboard, I was imagining, Lord, if, if I were to be looking at it, I would think about, oh, this is what a data's now the higher order thinking kicks in. How do I solve the problem? Was it just looking at it because getting access to all the information, it was extremely impressive and very, very inspirational to see your journey together. So thank you and I'd welcome, one or two questions from the audience, please. Patricia, you can

Speaker 2 57:19

use a mic at the back, if you don't mind, Patricia. Well, thank you for your presentation. And I was wondering about this gap between adoption people, organizations, adoption, and you said that trying to help your people to adopt this technology, but what we see is that everybody's using it. So how. Gap, works in the sense that some people are using this technically the whole day. And how, how is this your approach different from what people are currently doing? Is more focused, uh, special tools, uh, this. Something.

Speaker 4 58:09

Yeah. So, so in the, in the first place, it is, you know, o one of the things that I'm worried about is'cause I'm also responsible for data protection, data privacy, or all sorts of like, uh, our co top company risks when it comes to data. One of the things I'm really worried about is just open slither on AI tools that are out there. That means our data gets leaked or, or we introduce, you know, some AI tool in inside our firewalls that, that cause causes other problems for us. So, we. We, we, we strike the balance between blocking some of, blocking a whole lot of things that aren't, aren't ones that we've endorsed, but then also getting the right tools into the hands of the right people. So we're, we're working really hard with, and we gave those two examples of store managers and, and us Telstra for access to knowledge. Those two examples where they're really specific AI tools for really specific roles. And our goal is that we have role specific AI tools for every role at Telstra. And so making it really quite specific to the jobs that people do is how we're, how we're going after that because it, and then being able to plug that into our control plane and our other risk management assurance, you know, lifecycle management of the risk and lifecycle management of the AI tools. We can, we can govern that better as well. So, one, we're trying to get the right tools in the hands of the right people for their jobs rather than them just. Trying to navigate it themselves. and two, that then gives us a better, a better way of ma being able to manage the risk and the cost of them all as well.

Speaker 6 59:40

Yes, please. Jay Gonzaga University. Thank you for presentation. um, you at the end you talk about, uh, just the sale and can you share a little bit of how you measure that, what might be, might, might, works, what didn't work as well, and how you adjust the sale and the tool. You mentioned, uh, the more of top down button up do you adjust the cell to, to, can change a little bit of the feature of the tools. Can you share some experience how measurement lead to adjusting sales.

Speaker 4 1:00:06

Yeah. So, the, the US Telstra one and chat, GPT was released to the world in Decem in December, which is halfway through our annual cycle. So all of our investment was locked in for the year, um, because of the way our annual investment cycle worked. and so, but it looked like generative AI was, there was, there was some opportunities sitting in there, right. And so, we worked then to like, how would we bring that capability into our work? and, and we did it super fast because, uh, a, a timeline at Telstra, we, we got that, asked Telstra the first iteration of that in like up and running in six weeks. Wow. Which is amazing for Telstra, it's like years to be able to do something like that. But but the reason that we're able to do it was that one, we, we took that conviction based kind of bet in the first place. There's something in this, let's get going on it. It was going after a specific pain point for people. So there was a lot more buy-in from, it wasn't just a tech people experimenting with something. The business really wanted it as well. and we'd done a lot of work on the foundation. So we actually had a lot of our data sitting on Azure. We had Azure machine learning factory. We, we tried to get machine learning factory up three times before we were actually successful. Like, it's not all easy, just migrate your data to Azure and it's all magic. Like there is a lot of work in getting that stuff set up, but because we had that foundation in place, we were able to go faster on it. So. And then, and then being able to tip more investment into that based on some good foundations that we had, was a way that we had adjusted the sales there. Um, agen AI is something that's, you know, that like, that has, that has come through as well, and there's a lot of appetite for it. We're trying to find the right way to make the right decisions on it. but then the other thing that, you know, the, the thing that enables us to flip the, like adjust quickly is that, you know, how do we measure value, make those conviction based decisions, what foundations do we have in place, but then being really clear on what we will and won't do with our data, what we will and won't do when it comes to responsible ai. One of the things that we're really careful and, and cybersecurity as well, like, you know, uh, I mean you are in the US you get access to everything a lot quicker than we do. We have to wait for some of these LLMs to be available in region because of data sovereignty rules in Australia. and so knowing what we, we can and can't do around that, and then being able to, you know, work within those guardrails to be able to go after different things quicker.

Speaker 3 1:02:24

Maybe a couple comments just to add to what Dale mentioned. You know, I think there are two important distinctions I would make is for you to adjust the sales, uh, that Dale talked about. You need to be listen, listening very carefully. You know, I think your ability to listen to the signals that are happening, you know, both in the market, in the technology as well as signals from within the, your own organization is extremely important. Otherwise, you might miss something there. You know, like the agent TKI was a clear view. You know, as part of our, our program, we have got a AI investment forum that helps us to constantly listen on a periodic basis with the right executives involved saying what's happening, you know, outside in so that it can inform either we are on the right track or we need to change the track. So that's one dimension to think about. The second is education. Constant education of the senior most leadership. Around AI is extremely important. Like Dale mentioned, you know, we were together for one full day on an offsite with the entire CEO leadership team to educate them on what's going on. And that actually helped us to adjust the eight initiatives that you saw. Couple of them actually got co-created in that forum because they were getting educated and it was a dialogue going on. And then we prioritize and reprioritize saying, out of the eight, these two we take out, and these two we add in. And you need to have the flexibility to, to incorporate. And then the last thing I would say is, you know, the way you, you know, create the incentive mechanism within the company, you know, wherein it is okay to actually try a few things and fail, rather than always trying to set up for saying it has to always be a successful outcome. Because in AI we don't know many, many answers, right? And it has to be a leap of faith. We learn from it and the culture and the incentive mechanism should incentivize that, right? So how do you create that, you know, at the top most level and cascade that down is also important. please join me in thanking our speakers

1 1:04:11

again. And we'll be soon splitting into different tracks as as if, check out your program and decide where you'd like to be. So thank you again so much, and thank you very much, Dale. Thank

1:04:24

you.