From Data Scientist to the C-Suite: Foundations, ROI, and Leading AI at Scale — Kevin Cassar, Chief Data & AI Officer at TalkTalk

0:00Dr. Raoul-Gabriel Urma:The needs of the business evolved, so it's important that the teams will continue to learn and adopt to try new techniques. And it will all for them. My teams feel very excited. One of the attributes to a world-winning team is that satiation, unsatiable satiation to keep learning. So that need to keep understanding and learning and trying new things is one characteristic that I've observed of high-performing teams.

0:30Kevin Cassar:Welcome to Data and AI Mastery, the podcast where we bring you cutting-edge insights, practical advice, and inspiring stories from the leaders shaping the future of data and AI across the globe. I'm your host, Raoul Gabriel-Urma, founder of Cambridge Spark, the leader in transformational data and AI upskilling, career development, and progression. In each episode, I will be diving into real-world case studies of companies harnessing the power AI to drive innovation, reduce costs, and create new business opportunities. So whether you are an aspiring data scientist, AI engineer, or seasoned executive, this show is designed to give you the tools and knowledge to stay ahead in a world where data is transforming every aspect of business.

1:15Kevin Cassar:Stay ahead, stay inspired, stay masterful. Welcome to Data and AI Mastery. Hello, Kevin. How are you? Hello Roland, very good, thank you. How about yourself? Really good. I mean, I'm so excited to have you on the podcast today. We've known each other for a few years, so finally we're making it happen.

1:37Dr. Raoul-Gabriel Urma:Yeah, absolutely, and thank you for giving me this opportunity.

1:45Kevin Cassar:So Kevin, you've had a phenomenal career, working at some very, very big names. Deloitte, FCA, AXA, TalkTalk. So let's unpack that a little bit, right? Because I think your story is super inspiring. You know, I've met you when you were doing a level seven data science and AI apprenticeship, you know, in the C-suite. So could you maybe tell us from everything you've gone through and you've seen, what are the top three skills that, you know, has achieved data so you see how critical knowing what you know now yeah absolutely um so the first one is the

2:25Dr. Raoul-Gabriel Urma:the culture and that bringing people along the journey with me as a data scientist when i was earlier in my career i didn't appreciate it as much so only when you start to see the value the true value that people are people right although we deal with data numbers and technology there is a big human element to it so that transition into look we need to speak to people we need to empower people we need to bring people along the journey and we need to listen to people's needs to address them i think that's if i had to talk to my 20 year old self i would say look start to build this relationship early on because this will really help you along with obviously everything else the second part and this is the community which is talking about i i've been talking about this you remember me as a as apprenticeship straight from the start let's get the foundations right And there's a more increased focus about getting the foundations right.

3:17Dr. Raoul-Gabriel Urma:So that's the way that we think about setting the foundations and getting the basis right to help us to accelerate is another good aspect. The third part is people talk about return on investment. I talk also about route and speed to value. I'm talking about people that financial side, which previously I was an equity analyst when I started my financial career. that really helped me to understand what are the value drivers, what are the cost drivers. Not only that, how can we measure return on investment and as well as speed to value. The longer it takes to get the return on investment, depends on how you measure return on investment, the more it becomes costly.

3:58Dr. Raoul-Gabriel Urma:So those are the three skills. Getting to know people, getting the foundations right, but equally well, how you get the data financial profitability lens to what we do.

4:07Kevin Cassar:Okay, that's super interesting. So if you don't mind, let's unpack that. a little bit because one question that comes to mind immediately you know you've had a very technical career so how important do you think it is to be technical when uh when you take a

4:25Dr. Raoul-Gabriel Urma:position like a chief data in air officer so that's a question that i hear coming in uh from a lot of people that are about the difference between the technical aspect and hands-on enjoying solving problems or solves, etc.

4:39Kevin Cassar:versus leading and managing to others.

4:42Dr. Raoul-Gabriel Urma:There's no right or wrong. So it depends on what excites people. Both are important and that's a position to the individual, but it's a trade-off that needs to be done. In terms of the question, having the technical background and go through the ranks really helped me. It helps me in different aspects. The first one is it helps me to get respect from my teams much quicker and respect is earned rather than given, right? So when my teams get stuck, if they throw a challenging problem at me and give them some ideas, that really helps to earn respect very quickly, that they know that I know what I'm talking about.

5:18Dr. Raoul-Gabriel Urma:And I've been there, done that, and can share some of the experience with them. Equally, it helps me to identify many some solutions that might be at the end. So having that technical route and being there, done that before, can look at some of those aspects and maybe shape the trajectory of where we are. And the third part is when I'm sitting with my fellow EXO colleagues, I can educate them a bit closer to what it means, obviously, without going into the technical details. So keeping the right balance in there is important. And last but not least, it really also helped me to translate to my teams what is the EXCOM conversation versus what does it mean for you guys.

5:57Dr. Raoul-Gabriel Urma:And that is important because it helps my team also understand if we use the joke the janitor put a put a man on the moon a famous phrase in there that really helps the data team to understand my role is important this is the model that are building or whatever the data product they're building this is the value that is adding to the organization seeing the clear alignment really helps to thrive and empower people to do their job.

6:28Kevin Cassar:I'd like to take you a little bit, maybe in your more recent experience in commercial companies, defining and selling ROI of data and AI initiatives can be a challenge. I think we can align that. We can support revenue growth. We can support cost savings and so on. But could you walk us through how do you think about it? how do you sell those initiatives to what level of depth would you need to go in uh you know when you pitch to the rest of the exco or the board yeah absolutely but if i had to take a step back

7:01Dr. Raoul-Gabriel Urma:and keep at high level it's again coming back to speaking the language of the different exco colleagues it's not an easy task to do because if you've got other six or seven exco colleagues that means you need to pitch the same idea in seven different languages it's it's similar to saying happy christmas in 10 or 15 different languages right so that's that's what it translates to and get the bias so understand what are the drivers of the cfo coo cro etc and pitch it from their point of view this is why we're doing this piece of work or why we're suggesting this piece of work and thinking about the value that you're driving most of the time the c-suite colleagues will look about the value and the risks and the trade-offs between the do so pitching the not the data models that we're building that's why they employ me to leave it up to me to make sure that we use the right techniques and we can trust the outputs but then it comes to why do we care why does it matter and can we trust the output so if you can answer those three questions then you will you will start to push the right levers so again why do we care why does it matter and can we trust the outcomes you mentioned that you know when you pitch an

8:08Kevin Cassar:initiative you have to think about why it matters so that's kind of like tapping to the pain the problem what we can get out of it so why does it matter in practice i guess that's the the roi the solution that you know one could believe in and then are we actually going to believe it you know is it going to be a possible outcome could you walk us through one initiatives you deliver in the past that kind of fits this uh this framework yeah absolutely so when i was at oxa it's a health

8:36Dr. Raoul-Gabriel Urma:insurance company so the aspect is to give people access to health um private private health sector So that's what was one of the mandates of the insurance sector. Like data scientists, there are good data scientists and there are average data scientists. It's the same with every practice in there. So the mandate that we had, one of the mandate was to triage the number of claims that come to us to identify where can we give help, more help to other people that need it most. So some cases are not the same as others. and in some instances, some cases would need better support from inside doctors or physiotherapists, etc.

9:14Dr. Raoul-Gabriel Urma:Which as you can imagine, you don't have enough resources to give all this additional support to tailor the treatment to everyone. So you need to have a triage problem. So the business case that we had in there is how can we build up AI solution to help us triage between the cases, which are the cases that will require more support than others. And then how can we help those individuals? And the why do we care for this reason is twofold. So we can give customers better outcomes. So people could get better tailored procedures or medical care in order to get healthier quicker. But equally, well, that's on the outside.

9:55Dr. Raoul-Gabriel Urma:On the inside is how can we make the best allocation for the very scarce and expensive resources as well. So it's a win-win situation. and when you present it from that aspect we can give better customer satisfaction help us to achieve our mandate that we need to give better health to outcome to customers but equal to help us to improve our operational efficiency and manage the cost which ultimately can lead to lower our prices and better competition so that's the pitch why do we care and then it comes to can we trust the data so that's when we build models and there's different methods and techniques how you profile the data make sure the data coming in and out of the models that you build is what it is there's methods you know better than i do um how the models you can make a model explainability interpretability model performance metrics and what have you and let's then present that information digestible manner to whether it's easy to eat colleagues to give them comfort that we can trust the data that is coming out of our algorithms or models that we build that's a great

10:53Kevin Cassar:example. So in such a project, who would be the sponsor and what level of update would you need to give throughout the delivery of this initiative?

11:06Dr. Raoul-Gabriel Urma:Yeah, absolutely. In this case, it was the chief operating officer in conjunction with this CDO that sponsored it because obviously there's a data and there's operation aspect. So it was two people that co-sponsored that piece of work. And this brings in to your second question how do we ensure that the business comes along the journey with us and we tailored when when i set up this project i tailored at the time the concept of having multidisciplinary squads not just from a data perspective but including also the business colleagues the risk colleagues as well as obviously data scientists but from engineers

11:39Kevin Cassar:data engineers and and what have you and then and business spoke really highly about bringing them

11:44Dr. Raoul-Gabriel Urma:at the end of each sprint we worked in an agile methodology so at the end of each sprint we did a sprint demo where we presented what we set out to do in this initial sprint what we achieved at the end of the sprint and if there was a demo of the products that we delivered we we shared it with business and at the end of each print we asked the question is this aligned to your expectations and business felt very comfortable then to understand yeah we understand how the model is built we are completely aligned that the data product is fit for our needs and the business felt much more comfortable in owning the data products using the data products and deliver and drive the value that they committed to so bringing the business along the journey in an actual way obviously not day-to-day but at the end of each sprint was really key and instrumental in working collaboratively to develop data products that are fit for the business need to drive the value

12:40Kevin Cassar: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. You know, clearly the world is moving fast right now. There's, you know, new techniques, new models. there's a different way of think about use case and solution for them right so i want to take you back to a few years ago you did you know the level seven apprenticeship with cambridge spark so can you walk us through in your view how important is it to to upskill and what was your experience

13:24Dr. Raoul-Gabriel Urma:like yeah absolutely and that's a topic which comes up every day that you know what i learned i think it was about five or so years ago now the world has changed so much so the foundation stayed the same but gen ai have have really took off um since then so it's important to keep up up to date both myself personally so you know attending conferences listening to other colleagues reading publications podcast etc but equally well my team how can i ring fans time for them to stay up to date because otherwise there becomes a job a time when they won't be able to do their job equally well the needs of the business evolve so it's important that the teams will continue to learn and adopt to try new techniques and equal all for them it's it's my teams feel very excited one of the attributes to award-winning team is that satiation unsatiable satiation to keep learning right so that that needs to keep understanding and learning and trying new things is one characteristic that i've observed of high-performing teams in terms of my own personal experience.

14:32Dr. Raoul-Gabriel Urma:I say that having done the apprenticeship was a very powerful experience to not just go in the theory, but also help me to apply the theory to practice. And that was very fruitful to the organization that helped me to do the apprenticeship. Like together with other people, we set up a whole department through the algorithms that we developed as part of the apprenticeship scheme. It was a risky project. So I think if it wasn't for Cambridge Park to help me really the apprenticeship scheme was that project would have never got off the ground because it was high risk high reward but here we are now we can talk about the success of the story so it's important for people i'd say to keep an open mind invest in themselves it is not easy obviously if people are working and learning it does take a bit of input but it does pay back big dividends

15:18Kevin Cassar:yeah and you won the apprentice of the year award with bcs the british computer society so i mean phenomenal

15:27Dr. Raoul-Gabriel Urma:phenomenal result Kevin yeah thank you and I think you are with me on that day you remember the surprise on my face and that surprise comes in because I did this because I enjoy the job right and that's the message that I give to people make sure that you enjoy what you do and if you enjoy what you do the rest will fall in place right so I always describe that my hobby and my job became the same thing um and you know I enjoy what I do and that would drive would drive of the passion to learn, to adopt, to try new things. So that's the key success for me personally, to have invested so heavily in my career.

16:02Kevin Cassar:Amazing. And yes, I do remember the night very well being there with you. Good times. So one question I often get asked by non-technical leaders, which I'd like to ask you, you know, obviously we get excited by innovation. Data and AI can be channeled towards new customer experience, new innovative thinking and so on at the same time especially your ceo looking for you know show me the money because savings hygienic incremental improvements right some operational contents how do you balance out especially you know when it comes to deliverable and your team this kind of like conflicting position where we have to allocate time to deliver incremental operation improvements at the same time because ai is so disruptive we have to think about innovation and new type of customer experience how do you approach this dilemma yeah absolutely and

16:59Dr. Raoul-Gabriel Urma:having lived in the health insurance sector and the telco sector where that's one of the similarity that the customer expectations are moving at a neck-breaking speed right so on one aspect they determine better service low lower prices and anything in between as well as the competition between different players in the market so that creates the environment which you which you mentioned there all that there is the trade-off between moving at pace but equally well thinking about the tech debt mountain that you can start in care so how do you keep that trade-off between building sustainable long-term products that are aligned to the business value okay so sometimes i see organization that they go after the shiny products and losing that north star what does it mean for us why do we care but equally well there's always a trade-off between speed to execution and the tech debt that you are building.

17:51Dr. Raoul-Gabriel Urma:And that's why it brings us back to the start of the conversation, that focusing on getting the foundations right is important, because getting the foundations right can then enable you to move up at pace with minimizing the tech debt pile that you have. So doing the early investment up front, that's why I said it will pay dividends later on, because it will enable you to move at pace without building a whole mountain of tech debt. At some point, you need to pay it, right? That's why it's called debt.

18:16Kevin Cassar:So what sort of foundations do you need? Like what's the ideal, you know, set up for you?

18:22Dr. Raoul-Gabriel Urma:Yeah, absolutely. So there's no silver bullet to this. So you need to get your data foundations right. So you need to make to have sure that the data is in a robust way in terms of quality, in terms of accessibility, in terms of governance, in terms of governance. The governance aspect needs to be put in place as well. So do we have robust governance process that we build models in an ethical, sustainable way that are fully compliant without exposing unnecessary harm, both the organization and the consumers? Equal the infrastructure that we have. So then we start to think, okay, do we have our right infrastructure?

19:00Dr. Raoul-Gabriel Urma:Do we have the right platform? Do we have the right setup that can allow us to move at pace? You need also the culture. So to move at pace, you need to be able to bring the people with you at the same pace on the journey. So without that foundational education part, can people understand why you're doing stuff, the risk that you couldn't care, how you mitigate the risk and the profit. That culture piece usually is the bit that slows you down a bit to convince people to come along the journey. And in Kuluwa, we thought about this goals. so we need we need our workforce to be able to build up data products that are robust in a sustainable way otherwise as you mentioned coding etc if the code is not robust build up to industry standards then you start to build up code degradation code decay very quickly very very soon so it is a multifaceted way that we need to look at and once we get the foundations right in place and the organization can execute at speed.

19:58Kevin Cassar:So to be a bit challenging on purpose, like how do you measure all of that? And how do you tell your CEO, all right, I've got the foundations in place, right? Because you think about data quality, governance, infrastructure, culture and skills, what sort of, you know, KPI as a measure can you show and say, look, we're not quite there yet, but you know, we're closed and now we're there. Or is it an, you know, ever ending sort of story? How do you think about that?

20:25Dr. Raoul-Gabriel Urma:Yeah, absolutely. And you mentioned about KPIs, and that's something which is important and something which brings five weeks into my role. That's exactly what I'm looking at. What is the maturity of our data products, for example? And that's when the skills come in, right? What does a good data product look like? The documentation that you need to have, the quality of the code, the end-to-end execution, the ML Ops, if you want to go down that path, the scorecards that you need to have, et cetera, et cetera. So that's the experience when you come and you start to define what does good look like.

20:58Dr. Raoul-Gabriel Urma:You make sure that you use an agreement with other colleagues because sometimes there are some trade-offs. There's obviously the North Star, but then there's also what is practically where we need to achieve and where to set the hurdle bar. And once you agree those, you follow with them. So in answer to your question, that's when the experience comes in. What does good look like on all those aspects? What does good data quality look like? So the missingness, the errors that you have if it's unstructured data, how does it look like in there? What metrics can we put in there to make sure that the unstructured data is in a good, robust way?

21:31Dr. Raoul-Gabriel Urma:And again, that comes to the individual's experience and to help shape those metrics in place.

21:36Kevin Cassar:Great, great. So it makes a lot of sense. So, you know, a maturity assessment, having experience around what good looks like and showing the journey to get to what good is.