How the AA Balances AI Innovation and Human Care at Scale with Nick Edwards and Harj Johal

0:00Let's just imagine, right, we've got a family who have broken down on the side of the M6. It's midnight and we need to get a patrol to that family. We don't quite know, you know, what pets they've got in the car. We don't know what the fault of the injuries, but we know there's a family in distress. If we use AI and technology to get that journey better and make it right and more predictive and dispatch the right resources with the right parts, that could be a game changer. It could be absolutely transformational. all. Hello and welcome back to the Data and AI Mastery podcast. I'm really excited about today's episode because once again we're doing something a little different and this time we're heading somewhere most of us have turned to at the side of a road at some point in our lives.

0:45That's right, we're joined by two leaders from the AA, a brand that has been keeping British drivers moving for over a century and is now in the middle of a remarkable transformation story. I'm joined by Nick Edwards, Group Chief Technology Officer, who is building the technology infrastructure that is redefining how a 120-year-old brand operates in a world of connected cars, generative AI, and rising customer expectations. And Harj Joal, Managing Director of Customer Operations, who leads the teams and the human experience behind every breakdown call, every app interaction and every moment when a member needs the AA most.

1:25Together, we explore what the AA's vision of digital convenience with human care really means in practice, what it takes to bring thousands of colleagues with you on an AI transformation journey and why some of the most interesting questions in our industry right now are being answered at the roadside in the UK. Whether you're leading a technology transformation, running customer operations at scale or simply curious about what responsible AI looks like inside a business that people generally trust with the safety, this episode is one you won't want to miss. Let's dive in.

2:00Hey Nick, hey Harj, how are you guys doing? Hey Ro. Very good, very good. Great to have you on the Data and AI Mastery podcast today. Let's kick it off. So hey Nick, you joined the AI in 2022 from Lloyd's Banking Group. You've been recognized as a Global Chief Digital Officer, 100 Awards. that's pretty sexy. You've described your motivation as curiosity and a desire to solve problems that generally matter to people. So, I'd love to hear from you what problem at the AA first grabbed you. What's the opportunity? Well, many people know the AA. It's a really visible business in the UK. It's been around 120 years.

2:33First problem was the core of the business. Gentleman to my left runs our contact centres. Most of our calls were coming in for breakdown reporting to the contact centre. We wanted to build out a digital version of that, get one of times so first problem was really rooted in the the core of the business which is like when I need you how do I get you and how do I avoid having to phone unnecessarily so we we built a whole new breakdown reporting suite that was the first problem we got to work on amazing and I can imagine quite a journey to deep dive into in a moment and Harj you're a legend in the industry you've led teams of around 4 ,000 people across three continents so that's a massive scale so I'd love to hear from you what brought you to the AA and what made the opportunity so compelling?

3:15Yeah I think as Nick said firstly the brand is so recognizable you can't ignore it when the AA comes knocking. I think the second thing that really appealed to me at the opportunity was I had no predecessor there was no overall contact center operation there was lots of siloed operations so they were looking for somebody with domain experience to come in and really create a center of excellence around contact centers and clearly that's not always an opportunity that's presentable particularly when it's with the brand and having spoken to Jakob the CEO it was very clear the vision he had for the future years and it was delighted to come on board.

3:50Amazing so I guess to continue hard so you know Nick and you you guys work together clearly our partner you know you're in charge of the frontline business technology infrastructure and platform behind it can you walk us through a little bit how do you to collaborate? Yeah, Nick joined about three months after I did and we hit it off, you know, instantly from a personal level. I think it's really important because typically in contact centres or functions like ours, there's a level of tension between contact centres and the CDO department because typically you think the digital function is there to try to remove volume and ultimately roles or capability from the contact centre.

4:31So therefore, there can be an unhealthy tension. We've had the complete opposite from day one. We've got on brilliantly well. Our team see that and they say that this is a safe space and this is something that the AI has cracked in a way that other functions have struggled, which is we're trying to do ultimately what's the right thing for the customer. Reduce the effort and make it easier for them to deal with us. Make that more 24 hours than only the contact center times but Nick has delivered so much technology and capability that's actually made the roles of my staff a lot easier so that's why we embrace and we really collaborate with the digital function amazing so Nick is there like a joint like measure or kpi that enables this collaboration like what's the the secret to make this work so well i think hodge hit it off it's a bit like any technology you know relationships still matter so the fact that we get on really well we've got a common vision about the sort of a we want to create for customers and we're not myopic about our own worlds but massively helps but beyond that we've been given freedom to develop a joint set of KPIs a joint set of OKRs that move the business forward like also it helps that we've got a contact function that is genuinely interesting technology is on the front floor is pressing for better tech for colleagues better tech for customers so in a way we sort of push each other which I think actually is driven like good advancement for the group.

5:53So what would you say, you know, Harj and your team are so passionate about adopting technology? Like, why is that the case? Mainly because they are so passionate about customers and they want customers to get outcomes that, like I say, make insurance easier for them or make breakdown easier for them. But in a way that they also realise that there's something in it for them to be able to say yes to customers. And that's what they always want. and that's what I'll need. Many of my staff have worked in the organisation 5, 10, 15, 20 years and have actually had very difficult conversations and embarrassing conversations if you look back more than five years where they couldn't do the basic things.

6:33Dealing with fax and dealing with emails as far as the AA went back then and that is very difficult when you're trying to show a customer that they're dealing with a real outstanding organisation. Nick and team have come in and they've delivered some amazing technology and while that's digital technology It's been embraced by the whole organisation. And we're seeing that across many of our KPIs and metrics now are at the best level they've ever been in the AA. And that's due to the collaboration where my team understand that we are ultimately the people who have to bring that to life. We've got to embrace the technology.

7:05We've got to promote it with customers to say next time it may be better if you try to do that yourself. And we'll send them a link for digital rather than think, no, I only want you to deal with the channel that I work on. they understand the value for the organization rather than just one function. So Nick, as group chief technology officer, clearly you want to encourage technology adoption, accelerate business outcomes, but surely it must not be that easy. There must be some resistance in the business. So I guess I'm curious, how do you two manage to make it work? It's great to hear a passion for customer outcome, but surely there must be internal resistance.

7:43So how do you overcome that? I genuinely, I mean, I was in Lloyds Banking Group for 20 years before here, where you did face into pockets. I've got to say that AA is a pretty remarkable place. It's got a culture of can do. It always has. It's very operational. Every single day we'll deal with about 11 ,000 breakdowns. We don't quite know where they'll be or what the thought will be, but it's a pretty operational business. And therefore, the teams are used to finding creative ways to deal with problems. You know, weather. This country is so dependent on the weather. you know a slight dip in temperature or increasing rain will dramatically change the landscape so that that business genuinely saying has a very can-do attitude so very few pockets of resistance as we come on later even with AI the the appetite to adopt it was higher than the business's ability to scale it initially so there is something unique about this business where colleagues grab it want to run with it now what we've got to do is make sure we've got the right thinking the right guard rails the right measurement framework that's our job but the colleagues on the ground I find genuinely sort of welcoming technology across the business amazing so hard as a leader how do you support that culture of innovation and experimentation i think you've hit the nail on the head rail it's culture that's what it's all about culture it's for people to understand that as an organization we can't stand still we have to move forward and culturally they've got to play a massive part in that you know we're a business all around people you know we serve thousands of people every day and some of them in the most vulnerable situations and the human voice in that is always going to play a massive part but you've got to be you know we say digital convenience with human care and because that's the way we like to couple the two areas together there's huge demographics of the uk that won't want to speak to people they want to do it digitally they want to use technology and we've got to be there for them as well but on the cultural side we need the whole organization to move forward.

9:32We adopt an always ahead mantra and mission for the AI and that's what we're trying to promote. We've done that successfully over the last three, four years working with Nick and the digital team. And now we're at the cusp of another step forward using AI. But again, that will only be succeeded if culturally we take the organization with us. Amazing. So Nick, I guess, what are some initiatives that other leaders in other organizations can think about regarding the culture of the organization and helping people think about AI and other technologies? Yeah. I mean, as I said to you a little bit earlier, on AI specifically, we'd had a really intensive three years of digitizing the group.

10:13The group was behind, frankly, three years ago where it needed to be from digitization capability. And there's quite a trodden path on that. And lots of colleagues use that stuff in their everyday lives. They'll use apps, they'll use web portals, et cetera. or are we a bit behind the curve on that? From an AI perspective, I think that actually helped us because as much as we were catching up, we had all of these colleagues on this journey with us to digitize the group, build better journeys, be really intolerant of poor experience, really measure failure demand and get the feedback loops working between the channels.

10:44AI has sort of come onto the landscape in the last 18 months in a way that says, it's made us rethink about, do we continue to build down that digital path or is there a better solution where we don't just digitize, but we fully automate or we fully deploy AI to contain that journey that we was on a path to digitize. So I think in a way, the two portfolios have sort of merged and combined over the last, particularly the last year, to have a combined digital and AI roadmap for the group. That's fascinating. It sounds like being on this journey over the last couple of years has really enabled you to make the most out of what's coming now as well.

11:19So I guess, Harj, you have such a big team, right? What do you look for in people so that the transformation is successful? Is there specific skills or mindset that you look for that helps? I think there's the obvious things of people who have got a level of courage. They've got hard work ethic. But I think the quality that always stands for me is the curiosity one. People always want to pull on that string to think, but what if this or what if that? So they don't take the answer straight away. they'll go away they'll inquire they'll play they'll be curious with it and I think with AI that's the sort of mindsets we want because it's not a technology that's been around readily for 10 20 years so you can use it and make advantage very quickly if you've got the hunger and desire and the curiosity to really want to do that so that's what we're really saying to people you know we've been delighted with the work with Cambridge Spark and the transformation we're doing with all our senior leaders and I was at you know cohort one at Cambridge only last week and in the four months from the two visits the amount of progression our cohorts had made was phenomenal and to see some of the presentations I shared it with the executive team at our weekly meeting last week I was blown away with how far people had come in four months in their thinking and they were doing demos of live activity you know within four months and I spoke to them at the end of the day and said, some of you probably shocked yourselves by how far you've come.

12:49And they were, yeah, it is. But I could see that there was a real belief and confidence in what we'd work partly with yourselves and your organization, that if in four months you can achieve that much while people are still doing the day job, imagine when you really create this center of excellence in Nick's world and really get the technology flowing into the whole business, what the potential could be. Yeah, 100%. How does that make you feel, Nick? Because it feels like it's really special times. Yeah, it does. It's felt really good. I mean, we clocked really early on that this technology was going to be a little bit different.

13:23The AI has been through every wave of technology over the last 120 years, right? So you can imagine the path this business has been on. If you look back at our heritage, it's fascinating. We used to run this business with a whole bunch of cartographers building maps and all that sort of stuff. Very different business today. But it became really clear to us like a year or so ago that although there was a desire, a lot of our didn't quite know what to do and this AI you know you're being hit with it every day new technologies emerging new models emerging so we really thought for our leaders skills was going to be important like how do we get everybody to understand the way to reimagine the way we do things in the business so part of that was you know working with Cambridge Spark and building that program for our leaders secondly tooling you know we had some choices to make on tooling for our colleagues and what we found really early on was that giving colleagues the tools they are using outside of work had a massively disproportionate effect if they could use the same tools inside of work so that's why we decided actually to to buy 1500 chat gpt enterprise licenses deploy them securely get all of the guard works the guard rails and frameworks right so that colleagues could use them safely with all of our connectors and confidential information but like that combination of giving everybody a set of skills giving them some space and time to work with colleagues they don't normally work with to practice skills in a safe way and then a bunch of tooling with some skills and upskilling how to use it as that like a hard shed within four or five months we've just seen the water level rise across the organization really quickly into the space now we've got these cohorts i'm on cohort two we've got a bunch of colleagues on cohort three we've got 50 more use cases that have entered the system many of their way many of them are going to make their way to production pretty shortly yeah fascinating well i'd love to continue the conversation on the tooling aspect right so you've deployed a lot of licenses can Can you walk us through the strategic rationale for deciding what tool to adopt versus maybe supporting many tools and how do you make that decision?

15:18We hired a few people at the centre to come and help us with that thinking. So not a huge team by any stretch of the management, but a few people have actually been around this, have done it elsewhere to come help us. Two, we did deliberately pick a set of tooling and frameworks that was not irreversible. give us flexibility the market's going to change quickly over the next couple of years we know that so yes we chose chat gpt we've also got a bunch of developers using clawed code we've got co-pilot licenses so we've got a smorgasbord of tooling available in the group all deployed securely so that colleagues can use it with confidence that all of that's been taken care of and we'll keep our options open you know we keep really flexible on tooling to make sure that colleagues have the right tools to do their job of course you need some governance to make sure you don't end up in in a world where you've just got like, you know, a plethora of tools that are ungoverned.

16:08So we've got some structure, some thinking at the center, really effective technology review board we put in place actually to appraise tools, understand the costs, understand the implementation. But beyond that, we also realized like, you can't be too dogmatic on this. You've got to keep your mind open, experiment, and then just keep fixing forward. Yeah, great. So it's quite a fine balance to keep flexibility, but also empowering people to use the best tools to solve the job, which, Harsh, brings me to you, right? because you're like so close to the impact on customer and your team is literally at the front line there's all those tools that you know somehow have to be adopted and utilized so i'd love to get your perspective like what does good technology foundation looks like for you and what's it been practiced for the people on the ground yeah and i think we look at it from the outside of the business back in so how can we make dealing with the aa easier for customers and that stretches all the way to the thousands of patrols that drive around in their vehicles trying to fix vehicles.

17:09How can AI help them get someone quicker, fix the vehicle quicker and fix more of the vehicles? That's the ultimate ambition we have. That comes all the way back into the contact centres where I want my staff to be able to say yes to customers more when they phone about any query rather than two, three, four, five years ago and they'd say no we can't do that, now we can, yes we can do it and I want to give them more empowerment and permission culture at their fingertips so then there's no need to refer things off. I don't think many customers would say phoning a contact centre and being put on hold is something that they enjoy so that's something that we want to eradicate and make sure that we're seen as somebody that's easy to deal with, straightforward and just does things that you want and takes all the hassle out of life to get that done.

17:53We've got frontline colleagues have been given co-pilots and chat GPTs, they've created agents you know they've made agents to make their lives easier and then they've shared that across other peers to say look this is what I've built it's been approved it's got the the verification from the respective risk communities etc this is what it saved me would you like to try that and we've got this up springing from the bottom joining in with what Nick has said about the center of excellence and the top-down senior management so it's a real exciting place to work at the moment because everybody is talking this new language about AI, but we're doing it in a way that we've always got that customer in mind to say, that's the ultimate test point for us.

18:36How do we measure that we're seeing benefit through the customer's eyes rather than constantly looking internally? And how do you do that? How do you measure the benefits for the customers? We measure that many KPIs. We measure the most obvious KPIs you'd have in any sort of operation that I run about efficiency and productivity and cost. But for me, it's about complaint levels. If you're getting things right, you've got less complaints. We have. We have for the last two years. NPS is moving up. It's, again, at the highest level it's been for two years, etc. Satisfaction levels are moving the right way.

19:09But I also think it's really important that you see from the organisational staff point of view, our engagement scores are up. So you could argue that people would say, this is going to be seen as a threat. it's not they're happier working in the organization now than any point in the last five years so when you look at that all together you say well this has got to be working the right way because we're very very mindful of the impact on people we are openly transparent and honest with what we're doing and why we're doing it and why we think it's a benefit for people and also how we can upskill many people have worked in the business for like i said more than a decade two decades this is the first time they're getting real personal upskilling and and learning something new in their job and what that allows them to do is embrace and increase their own personal toolkit and the first time they may start looking for other internal roles rather than being pigeonholed in one job that they've done for for the last few years that's fascinating because it's such a valuable skill set that really encouraged maybe internal mobility 100 and that's what we want people to do to be able to do more and look wider across the business because that breeds collaboration then when you can do your job and know more about somebody else's you'll partner and you'll come together and if that is always looking to solve business problems that ultimately benefit a customer that's great momentum together yeah i like that uh seems to be a common thread here the north star is what's right for the customer 100 there's one other bit i'll add to out there.

20:35We hadn't quite designed at the start. When we deployed Enterprise GPT, we had a rollout program. In those first few weeks, we saw in the data some really early adopters, and it wasn't what you expected. It wasn't the technology function. It was in the operational functions, in finance, in HR. So actually, the next thing we did was rapidly spin up a champions program. How do we get 75 of the most engaged, galvanized people around the group who are showing really early adoption curiosity as hard said we've built 300 gpts in the first couple of months that wasn't expected so actually this champions network 75 colleagues are all levels not hierarchy thing across the group who are helping others who are the evangelists on the ground who are running brown bag sessions lunch and learn it's been so powerful and that hasn't been a top-down program that's been a bottom-up program these guys are super engaged they want to help others and i think it's a good way of actually understanding that in this you do need some top-down frameworks and all in for sure but the bottom-up excitement and turning that into something that colleagues can help each other with has been really powerful yeah well i really love this combination of top down and bottom up and it sounds like that's creating a lot of uh you know demand and uh new projects and tools and so on and you know it could be a challenge right so i'd love to ask you the question on you know how do you manage all this internal excitement and all this production of new projects and agents and so on in a way that still benefits the business is there like a framework to approach this i mean it's really hard we we definitely went in with a theory of let's not let a thousand flowers bloom in an uncontrolled fashion but on the other hand once colleagues start building gpts and having access to it's really hard to to control and you don't really want to over control it so we've we've been really selective like the group will make a number of calculated but big bets.

22:25So we're going to make some choices about where we deploy it. We know our business, we know what we do, and we know where there's huge opportunity to improve the business. And we're going to make some calculated bets in that space, which we've already started. But beyond that, we'll also have some freedom within the framework. We'll allow some... In fact, today we've just launched a Dragon's Den type process. We've put some money in a pot to allow teams to access that and do some low-code, no-code, maybe mid-code development within the teams. So a combination of both. So yeah, it's curated rather than controlled, but we sort of know the pockets of opportunity that we want to work on over the next 18 months.

23:00And they're really exciting. That's amazing. I hope you're enjoying today's conversation. If you're finding the insights useful, please do take a moment to subscribe to the Data and AI Mastery podcast and leave us a review on Apple Podcasts, Spotify, or YouTube. Every new follow helps us reach more people and shed incredible work being done by today's Data and AI Leader. All right, let's go back to the episode. So is there maybe a question for you, Harje? Is there like a framework to filter ideas that, you know, yes, they're going to be impactful versus it's probably not the right time or it's not going to be impactful?

23:35Yeah, the drag and den approach is open to everybody in the business. That's the first thing. But we'll probably control it at a directorate level first to say, OK, let's support and give everybody the best chance. What you said earlier is so important that everybody now is behind this and wants to do something. so you've got to have a level of control around it. You've got to make sure that happens. Otherwise, it can be organised chaos and everyone's doing that. And then suddenly you remember the day job's got to be done as well so everyone's off working with AI. So I think it's getting that balance right.

24:01So I think we've got a measured approach now to say, okay, let's be very transparent what people are working on. There's no point two or three teams trying to do the same thing. Let's come together and make it even more powerful, etc. So sharing of ideas is really important. Then it's a case of, okay, which bucket does your opportunity fall into? Is it productivity? Is it efficiency? Is it a growth thing? Is it pure innovation? And how can we support that? So we're really nurturing and supporting all the ideas before they turn up at this, say, a Dragon's Den event where you're pitching them for this real investment and to say, look, let's take this forward at organisation level.

24:37I said, Jakob, the CEO, launched that only today. And I've already had messages pinging on my phone from people saying, look, this is fantastic. Can we take our idea forward at heart? So it's already like I'm like, it's only started. Let's get the right people together. But that hasn't really happened with many things in the organization when you've launched it. You normally launch something, you've got to give it a push a couple of times and then people take to it. It just shows culturally, I think we've done well in how we've promoted AI and how we've upskilled and how we've partnered with people like yourself.

25:09And I think people now really excited about this is the next step that you can really come up with ideas and innovation that are going to be back to organization level. And again, that doesn't really happen everywhere at the level that we're promoting it. Yeah, fantastic. So to balance it out, Nick, fast forward a couple of years. What do you see are the big opportunities, but also the challenges that might be coming your way? If we get it right, the opportunities are really significant. If we get it wrong, it really is a really operational business, like I said earlier. Let's just imagine, right, we've got a family who have broken down on the side of the M6, it's midnight, and we need to get a patrol to that family.

25:55We don't quite know what pets they've got in the car, we don't know what the fault of the injuries, but we know there's a family in distress. If we use AI and technology to get that journey better and make it right and more predictive and dispatch the right resources with the right parts, that could be a game changer. It could be absolutely transformational we get it wrong it's equally bad for that family so at the end of our you know there's always the last mile of our journey is where our you know our shining knights in the fluorescent gear turn up to go and rescue someone so it really is making sure that we've got the roadmap to make that experience and cars are getting more connected than ever right so we can ingest data we can do all sorts of things in the future get it right it's an absolute game changer for us but we also know it's high stakes it's high risk these are really important moments of truth they don't happen very frequently and we've only got one chance to leave a lasted impression on that customer of our brand so like that is a good example of how we're thinking about the opportunities but also the risks and there are yeah i love it at the end of the day it's really thinking about the uh the end customer right and like seeing it through through their eyes and that's kind of like your guiding light in a way so that's that's really interesting And Harj, how do you balance out, you know, over the next couple of years, right, this idea of incremental improvements versus just reimagining, you know, how the customer operation experience looks like or even how the customer interacts with the air?

27:21So how do you balance those two together? I think for the contact center industry, I think this is probably going to be the biggest change for two, three decades, particularly my lifetime of working in the industry. I think we're on the cusp of something very, very special. And I mean that not only for customers, I mean that for people working in the industry. Firstly, I think the contact centres will always be needed and will always be present. However, I think the role of a contact centre individual is going to change significantly. I think the job is going to be upskilled to a level that they're going to become orchestration hubs that are data-led with AI technology that are single points of success for individuals and make a whole business operate by the fact that they can look across many domains, many touch points, many industries, and one individual has got the ability to access that much knowledge and that much capability that's never been seen before.

28:17That's agnostic of the back-end platforms they're using. I think they will all find a way to be linked to make sure somebody can do that. So you'll really have for the first time a cool one-stop shop because that example Nick gave about the family breaking down, That may still start with a phone call for the first 60 seconds, but then you might find that you've got that individual able to access a car, see the data, understand where the patrol is, link the two together, make sure that people are safe. If there is pets, et cetera, make sure we've got a concierge service set up. Is there a hotel nearby?

28:51Get that ready because these people won't be able to get home and we won't get the vehicle home because I can see from the access. so don't send a patrol van get a recovery truck going straight away because i've looked in the vehicle and that can't be fixed so we're gonna have to move them etc and that is all being orchestrated by one person who's owning that situation historically three four years ago that may be five six different individuals in the business owning their part of that going forward the technology that's been we're now looking at should be able to say that's one person's responsibility to own that end-to-end and there is nothing more customers love than knowing that somebody is looking after them and dealing with that whole issue they don't like it when people hand off put them on hold or transfer if they know they've got somebody who's looking after them at their most vulnerable moment that is why they deal with organizations and stay dealing with those organizations I think AI is going to allow us to upskill our people and then for transform the role of a contact center worker to a level that's never been seen before.

29:51Wow, that's a super compelling vision, right? Being able to empower that level of ownership around the customer outcomes through accessible data. Hey Nick, so what advice do you have for an executive that is looking to demonstrate ROI out of technology and AI? Where should they start? How should they think about it? Any recommendation? Well, I think all of us, no matter what role you do around the executive table, I'm in technology, hard runs ops, you have to lean in, you have to be curious. This is not something you can outsource or completely delegate to the technology function. It's as much a business transformation topic, a colleague transformation topic as it is a technology.

30:33I think that's a really important part. Every one of the GEC in our company is leaning in. They are all going through the same learning with their colleagues in a very non-hierarchial fashion. point number two is i think keep an open mind about the technology it's really easy to be too skeptical at first and for sure there's hype right there's there's definitely hype in this space but in that it's about navigating and finding the genuine opportunities the partnerships the connections keep current so actually you know if you've been in digital around technology for the last 10 years you had to do that you had to eat current make connections attend the right seminars etc i don't think this is any different i think the difference is the pace at which it's moving and the fact that we've all not been using this in our day-to-day lives for that long.

31:12When did GPT come out? November 22. It's not been that long. We've all been using apps now for 15 or so years, right? So I think the familiarity of it, it can make you feel a little bit overwhelmed and daunted. And even, you know, the last 18 months, I've had points where I thought, I just cannot keep up with the pace at which the thing is moving. But don't be daunted by it. Stay curious. Stay connected. Work as a team. And do you know what? The bit I said earlier, you'll find some real pockets of innovation and surprising places you didn't expect to find it in the group. So keep an eye on the metrics, see who's using it, see who's posting on the internal chat networks because you'll find someone that you didn't expect doing stuff that's super cool, go talk to them.

31:48It's interesting because both of you said curiosity is like a key skill set in times like these that are moving so fast. As a follow-up to that, Nick, clearly a big investment in lots of licenses in the business, get the business on the journey. How do you think about ROI and P &L? Because in reality, a lot of leaders are trying to figure this out and they might not be seeing the returns yet. How do you think about it? Yeah, I think it's a great question. And I guess we've deployed sword and shield in some respects. We don't want to be too defensive, and we know we need to keep an open mind, and there will definitely be some learnings through investment where it doesn't quite pay off.

32:23So we're trying to make sure we stay disciplined on investment and don't just start spending money willy-nilly on it. But also, deploy the sword. We're confident that the technology can play very strongly but very strongly to some of our business cases. So, you know, there's a lot of demand for investment in the A. There always will be more demand than supply. We're applying our disciplined approach to how we allocate capital, but we're also realizing this technology is a little bit nascent. It's a little bit more nebulous in places, so have some room for innovation. But as I said earlier, this can't be a thousand flowers blooming and no benefit at the end of it.