Transforming Healthcare Through Data, AI & Human-Centred Design — Ming Tang, Chief Data and Analytics Officer, NHS England

0:01Everybody's busy these days. You want to be able to interact with the NHS like you do with your banking app. So we've got the NHS app to do some of that. We want to have a consistent experience. That means we need to bring everything about you as part of that journey.

0:20Welcome 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 of 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.

1:05Stay ahead, stay inspired, stay masterful. Welcome to Data and AI Mastery.

1:14Hi Ming, how are you? Good to see you. Good to see you. How are you, Raoul? I'm doing great. I'm so thrilled to have you on the show. You're an amazing leader, and I can't wait to have a really interesting conversation together.

1:34To kick us off, Ming, I'd love for you to maybe tell the audience a little bit what your role entails. So you're the Chief Digital and Information Officer at NHS England. Would love to, in a nutshell, what does it involve? yeah so i i've got two roles at the moment the digital thing is really bringing together digital data and analytics so that we can help transform the nhs um i think it's actually a real opportunity of privilege to bring those things together because previously um we had quite separate digital work and then data work and obviously to help transform the nhs we need to bring those things together to work in unison.

2:17So I've been given the opportunity to do that, which is fantastic. Amazing. I wonder, where does that passion come from? You know, this intersection of, on one side, technology and data, on the other side, you know, delivering real impact to people. Yeah, where do you think the passion comes from? The values of the NHS are fantastic. it's an absolute privilege to do things at a scale that we do as the NHS so now it's really how do we garner everybody and make sure that the experience for every patient is as good as it could be That's really inspiring to hear the combination of obviously leadership capability and technology but channeled into a personal passion to drive impact in the country so thank you Ming

3:14Well, I'd love to take you now to your journey, right? Like, you know, you clearly, through your transformation journey over the years, there must be important lessons that, you know, you've acquired along the way. What would be like maybe one lesson that you've learned from leading the tech transformation in the public, but also private sector previously? One of the key things, which sounds a really simple thing, it's actually not about the tech or whatever. It's actually just listening to what the problem is and being a bit more curious about what it is that's creating the problem because, I don't know, 60%, 70 % of the problem will not be technical.

4:00It will be about how human relationships work. So I think over the years I've learned a lot more about how we behave, how we learn in order to solve problems and then it becomes a team sport doesn't it because it you can't solve it as one person so I think combined with listening bringing the right people together and then having a bit of a technique so my consultancy background actually helps in kind of breaking down the problem trying to understand where the common patterns are and how do you then carefully bring orchestrate the right people into the room so you can have those diverse conversations both you know in health it'll be the frontline people it'll be you know clinicians as well as operations people administrative people understanding the pain points so that you can then solve the problem you know years ago we used to call it you know how do you how do you improve lean or whatever the process i think at the moment what we're trying to do in this space is how do you bring the best to bear to solve the problem as quick as you can and then technologies move so far so fast you can always apply that solution or some assets to solve those problems so i think it is it is listening it is really being curious about what the common problems are because you can apply solutions from one area to another if you if you look out for those patterns.

5:33And mostly, I think where the data comes in is the context of that issue. So bringing the right insights and the right data to solve that problem in the context because the 20 % everyone worries about is their context. Do you really understand what it's like to work in a theatre? Do you actually understand what it's like at discharge? Do you actually understand what it means for me in the front line every day when I'm banging my head in A &E? And that's the contextual thing. And that's the value of the data, because you can then say, based on those pain points, this is the process that we're trying to change.

6:10And here's the data to make it real for your context. That's really interesting, because I was listening to one of your recent keynote at the healthcare conference, I think it was around the summer. And you made a very good point, which I'd love for you to elaborate on. You made the point that we tend to focus perhaps on standalone services, you know, like primary care, secondary care, admission and kind of like look at optimizing them in a kind of decoupled manner, which is what we like in technology, you know, decoupling and efficiency. But in the eyes of the patient, you know, I should have exactly that context, right?

6:47Like what about me as a whole? What's my experience? How can you add value to me across all of the services? so can you elaborate why do you believe that's so important because i think that's quite a fundamental principle that applies in public and private sector right thinking about the customer first yeah so we're in the if you think about how data is collected in nhs it's been built around each service that people come to so in order to do a good experience we need to join those things up and we need to think about it as a journey right and that's what consumer goods companies do really well but they don't own all of it the the value we have and the opportunity we have is to really redesign our services in a way that is intuitive so you know we are we're all busy everybody's busy these days you want to be able to interact with the nhs like you do with your banking app right so we've got the NHS app to do some of that we want to have a consistent experience that means we need to bring everything about you as part of that journey and that's why we're thinking about the single patient record bringing your longitudinal piece around you and and then you know we want to make sure that by connecting that care we have the next best action for you so we we can bring some of that predictive um understanding and tracking where where things are going wrong for our administrative staff our operations staff across those different silos um so that's where the ftp comes in and then really what we're really trying to do is connect you know we're never going to have one humongous one organization that does everything thinking simultaneously but we do need to have those connection points so that the jobs that we're doing are connected and breaking down the silos is both how do you come together as multidisciplinary teams to support care for an individual in the community or in in the hospital but actually having everybody having the right information to do the right things is really important to that so breaking those silos down is just a natural thing if you want to follow the person around their treatment and it's a necessity because if a pharmacist doesn't know that you're having certain treatments that may you know they don't have access to your record trying to make them be aware of that is really important obviously also for the public you know i know patients that carry around massive files because they've got complex conditions because they have to repeat their story every time and what we're really trying to do is make sure there's a summary of their record so that that can be surfaced at the right time and they also know what's being said about them and they will also have access to better understanding of their care 100 percent now it feels like maybe from a implementation point of view there's a bit of a bit of a tension because I guess there's benefit on decoupling teams so you can move faster independently by minimizing interaction, kind of minimize inertia.

10:01So there's some benefits there, but the downside is you're preventing collaboration, which benefits the customer and in this case, adding patient value. So I do wonder what's the balance or what's a good organizational structure sure that you know people can can learn from it to kind of like balance out the need of the customer the patient versus have an efficient backhand to kind of like you know still make progress fast like what's your view on that so i think the trick is bringing all those multidisciplinary teams together to help solve the problem including the patients by the way um helps to solve the problem or identify where the pain points are then technically we can bring the right mix of people to start thinking about how we solved the problem.

10:50And then in the background, your back office bit is how do we stitch these things up? How do we create the backbones, the core services that are efficient and effective and can be created once but used by multiple services? So it's how do you create a set of microservices on very strong core platforms, which then allow you to have that flexibility of plug and play that's the ambition that we've got we're not quite there yet but that's what we're trying to do because the core capabilities the way i look at it is in most processes there there will be if you're curious about the steps at a high enough level a capability level they're very similar but the context as i said earlier is very different an appointment in any setting is the same but we've designed it differently because we've designed it in the context of that service.

11:46So some of these things is about abstracting that so that you can use the technology to feedback and interact in those systems. But from a design perspective, we want to simplify and actually not design what we're doing now, trying to imagine what it could be with the help of technology. Yeah, I love that. 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.

12:26All right, let's go back to the episode. So I guess on the topic you've mentioned changing heart, changing mind and, you know, embarking on this journey requires big transformation. you know we talked about the organization level um what do you measure like i guess you know on that journey what's worth measuring what's important we are changing these processes for the better good right so we need to have a business owner who actually says this is how you're helping me change my business does that make sense in the commercial sense that's very common having the business owner for all the change then allows you to say whether the technology and has actually helped you deliver that.

13:10And then the metrics become very part of the business metrics, whatever you're trying to solve. And it becomes very concrete. When it becomes adoption, if we start measuring the metrics on just adoption or percentage of uptake, that's less value. that's in there's an image is an important metric for the tech guys it's not transformative of the business so if my if it's speed to market if it's tight you know reduction in my waiting list that's to me a tangible target how has technology enabled that it brings us into a complication of what's the contribution of technology to that problem and are there process changes that you do which is i welcome that discussion because that means we've got into the nub of it we're not just putting in some technology for the sake of it we're actually trying to change something right and together it's about a partnership and how much does technology actually contribute because it won't all be technology the whole point of it is we're actually doing some of those changes in public sector we have to go through you know a five six case business case it's very formalized it's very methodical and then you say you know what's the percentage productivity that you've gained which is quite abstract in real terms but so measuring some sentinel changes so has the patient experience changed measuring some really concrete business changes is important and then the the pace at which we've done this is also important so it's a you'll end up with a balanced scorecard if you do it properly great great so this makes a of sense.

14:53Now, how does it work when you have those more sort of a horizontal change, right? So for example, the federated data platform, which is something that's a bit more cross-cutting. So I'd love to get your perspective around how do you make the case or how do you show ROI for the sort of maybe bigger horizontal sort of transformations? So what the federated platform does is allows us to have a common data platform with the new technologies to do interactions, APIs, but also workflow. So when you redesign a process, some of it you want to enable through workflow. And what we think about, the way I think about it is in the logistics of care, the coordination of that multidisciplinary team and the coordination of follow-ups and all those things you have a number of you have a tech stack that allows you to work between them so the NHS app single patient record FTP they all work in conjunction but without the federated data platform we wouldn't get as much connectivity as we would do with those just those two things it would make designing the other things on top quite hard so there is the opportunity in FTP now when we made the case to treasury we chose five use cases because we wanted to focus and we could we could gather the benefits sufficiently against those five use cases in order to make the investment case so the return on investments was based on the business change and also um the use of the federated data platform to create new applications that would increase theater utilization it would look at population health it would look at um coordinating care and um support a number of other things so logistics of care in screening and vaccinations and the supply chain so there were five quite distinct use cases that we fit that business case around and that's my advice really if you if you're if you're using technology has a reusable capability you'd be very clear on what the use cases are because then that allows you to illustrate value and then i think what we've done is we've gone into that and tracked value in each of those applications that we've developed we have a benefits framework that actually tracks benefits has that application is rolled out amazing super helpful so speaking of core capability you know this is the data and ai mastery show so we kind of talk a bit about data i need to take you to the ai side as well and maybe let's start kind of high level right like how do you think about ai as a core capability in this in this context i think we've been doing some careful thinking about ai i mean the government's had lots of you know how can we accelerate adoption of ai i think in health we've actually embraced it quite a lot so because we've had the federated data platform there are LLMs already embedded in that platform so that's allowed us to play and test so we have a discharge summary tool that's been developed using AI which basically allows summarization of everything that's happened to a person to get them ready for discharge and then making that summary available for doctors to check and then use, and that goes into the record and it gets communicated to the patients and the GPs.

18:31So that's been really a good test. Because of the nature of where we are, we have to get that through regulation because there's some, you know, we've got to work that through with MHRA, our regulator on AI. So that's one side of it. The other side of it, you know, we've got ambient voice technology that's actually really popular with clinicians because that voice-to-text, scribing, helps them take time away from the administrative task and it frees them up to actually look at the patients and take time with the patients. So that's very popular. We're just working through what does that mean. I think some of the other things that we're looking at now is really building on the data that we already have.

19:18So we're playing around with, you know, a assistive AI piece, which is really helping our analysts do some of their work more effectively. So based on the data platform that we have, how do they ask the data we've got in the ontology to help them create charts and other things? So, again, that will be helpful. and the longer term part of that will be asking any questions so you know most most people have put a chat bot on top of their knowledge base we'll be doing that with some of our data in the future we need to test it we need to make sure that you know those models are are not moving and they're they're consistent so using some of the foundational models to build some of that is is kind of the focus for the NHS.

20:13It's exciting times. You know, you can really see how AI could help people have much more of an intuitive interaction with the technology. And the data, with masses of data, it's too much for any one person to be able to get across. So, you know, structuring the data catalogs, making sure the data profiles are right, the quality of the data is really important if you're going to apply AI on top. That's great. So in that context, do you have any advice or guidance for leaders out there that are looking at supporting AI adoption to solve real business problems? What should they think about or what sort of mindset would you recommend?

20:59I think the mindset I would recommend is to be curious. There's different types of AI, so it's not one size fits all. So understanding your problems, breaking it down will then allow you to, you know, is it summarization? Is it scribing? Is it a chat box? Is it assistive AI? You know, if you're thinking about what's the assistive part of the process, I think there are some data foundations you have to have in place before you play too much. but the exciting thing with AI encoding and data is that you could also use it for some of those back office things like getting your data in a better position you know we're starting to explore using AI to look at how some of our legacy systems how can we use it so that the code is raised so that we can then use it more you can apply APIs to it etc so there are different uses so I think it's exploring where your biggest bang for your buck is where you're going to get excitement from the business because some of this is about people fearful how do you bring people along with you so is that where's your sweet spot so is it about turning leadership into being advocates is it about getting your teams to buy into it and then how do you make sure you've selected the right type of AI in that context and then have have a go there's no harm really to have a go and then seeing so long as you put some guardrails around it and that you you really do test what you're this thing that you're trying to solve so being a bit more struck is again it's structured and free flow trying to get those two things in the balance so that you you can apply AI to real life problems based on the right type of model and learning great yeah i love that

23:01thank you meng can i take you to a quick fire round of questions just the short questions super um this is actually a big one so if you had a magic wand what would be the number one issue you wish you could solve tomorrow i wish i could solve the waiting list problem i don't think we're quite there yet. Thank you. And then a contrarian view that you have specifically in the data and AI world. Probably that it's not all about the tech. It's actually about mindset. Yeah. Because if you can apply the mindset, the tech's there. If you focus on the tech, you miss the trick in terms of delivering value.

23:42I agree with you. Next is a bit more personal. What was your favorite subject at school? Biology. Biology. Ah, fascinating. Next question is, what's your favorite programming language? Oh, I don't really have one. Okay. English? Today, I guess English counts. Yeah, with AI, English counts, isn't it? So I think if we make it, I mean, obviously, So you need code underneath it all. But we are getting to the stage where we want to be native, don't we? Actually, whatever language we're speaking, we should be able to apply the tech to. Because that will be the intuitive way to interact with technology.

24:30So I don't really have a coding base. You know, probably Python if you really want to push me. And last question. What's your favorite music genre? I think I'm still stuck in the 90s.

24:48No, I don't. I like all sorts of music, actually. You know, some, probably something you can sing along to. That's fantastic, Ming. Thank you for your time today. It's been a real pleasure to talk to you. Lovely to speak to you too.

25:12I really enjoyed my conversation with Ming today. What a really inspiring leader. There were so many cool nuggets out of our conversation. The first one is, as technology leaders, we tend to get excited about technology. We think about how to make our services more efficient, how to introduce decoupling in architecture and so on. You know, this is great. But at the end of the day, it may not be moving the needle for your customer. in this case you know let's think hard about what's the value we're adding to the patient how can we reimagine experience how can we make the life better so starting with the customer and the user first and technology after technology actually is an enabler so as leader let's focus on the customer first the other really cool nugget was i asked her hey what's the mindset that you know you'd recommend leaders to have in this context where ai is moving really fast and she mentioned curiosity experimentation learning you know those all kind of characteristics that you really need to embrace as a leader in order to support your organization with the adoption but also for yourself to understand what's the art of the possible and get the business on that journey right and she made a good point get the business excited focus on what's exciting for the business so thank you everybody great episode today Thank you for tuning into this episode of Data and AI Mastery.

26:38If you found value in today's discussion, make sure to subscribe so you never miss an insight from the leaders driving the future of data and AI. And if you're a data and AI leader looking to upskill your workforce with the fundamental data and AI skills to transform your business, Cambridge Spark is here to guide you every step of the way. Be sure to reach out to us on LinkedIn or on our website, cambridgespark.com. Until then, be sure to keep pushing the boundaries of what's possible with data. And remember, mastery comes with continued learning and action. Until next time, stay ahead, stay inspired and stay masterful.