Innovating Social Care: AI and Digital Transformation

Hello, welcome, my name is Mark Jackson from the Intentive and Learning Academy team at the University of South Wales. Welcome to the latest edition of our Change Management podcast. I'm joined today by Shameen Smittel, who is a service designer at Social Care Wales and a recent graduate over Master's Program in Leading Digital Transformation. Welcome Shameen. - Hey. - Good to meet you. So in this episode we're going to look at your final major project and we'll let you tell us a bit more about the details of that. I think it will be useful for everybody if we could just outline what Social Care Wales is first of all and give us a breadth of the diversity of the challenges that you're dealing with and the kind of sector and all of the things that you can come across as challenges. Sure. So yeah, Social Care Wales is the Social Care regulator within Wales. That means that we register social workers, domicillary care workers, care home workers and managers. It's quite a diverse range. It's about 66,000 people. So they all need to register with us so it legally within Wales, within those job roles and sectors. So it's quite an important thing for somebody who is looking to, you know, workers, a care worker to have to register with us. They pay a fee. It's quite an important part and then every three years they demonstrate that they're still good to stay on the register and work within the sector through demonstrating their CPT. We carry out some other things. Just keeping us up to date really with all the things that they've learned within a three year period. So really we find to see our role as kind of offering learning, development information, research, data and innovation. And we know that when our registration part of our service kind of falls down, it can be people vulnerable, it can leave people out of work. So we really were always working to try and keep that as slick a process as possible for end users and for the workers in the workforce. That's brilliant. And 66,000 users is a significant volume of users. So in terms of the diversity, we've got people coming in from different backgrounds. Some people are obviously Welsh. Some people moved into the country recently. People are working at all different levels, I guess, with different languages and different needs and all of them working to support social care across the country. Yes, it's really, as you mentioned, read diverse. There's some people that are educated at a master's level. They have placed graduate degrees. All there are people who have done workplace learning. So that means you have, as you said, diversity have understanding and learning different languages. We know that the sector has got a lot of languages and also different digital competence as well. So we're really trying to cater for quite a lot. So that brings us to your role as a service designer. So what's a service designer? A good question. So really what we do is we work to really understand the people who access the service. So who need our service, they might be new people or existing people and find ways really to make that service work for them and also deliver, I guess, help us meet our business goals and Welsh Government priorities. So we're doing a lot of research, analysing data and a lot of creativity. But within our organisation we're really focused on really understanding the behaviours of people who use our services to help inform how we design things. Brilliant. Okay. So we could probably come back to that later on. So that really leads nicely to your master's project. So for the context for our listeners, Jamehni's just completed the MSC in leading digital transformation with us, which she did alongside her professional role. So really busy professional life as well as this additional challenge of study alongside alongside your work. The way we do the course we kind of integrate those projects, so students will identify, student, Jameh, will identify a project that they'd like to explore in much more depth and then work and collaborate with author employer and with our academic team to develop, make potentially an intervention and develop a project through to its fruition if we can. So that's where we met last week and I was blown away by the depth of your understanding of the project. So I think for our listeners it would be incredibly useful just to give us an outline of the project and then we're going to take you through in the same ways we did last week. Yeah. So we're, I guess, undertaking a transformation as part of the registration service. So my projectors were small part really of that transformation where we were looking at, we did it taken a review into the service. So my project was focused on how we might be able to use AI to really help us solve some of these problems. So we're looking around how people within the organization access work instructions and also looking at how we might be able to use AI as part of our broader customer service offering. So we have lots of information here and it's really about making a lot of that really accessible and easy to digest and understand within the organization. As I mentioned, they're registering a lot of people dealing with a lot of application forms. So we're just really trying to make that as easy as possible. So for an example, somebody might choose to join a new organization as an employee. Often I'm conscious lots of postgraduate students might register with social care wells in terms of some additional work that they might be involved in. They have to by law register and then they obviously then have access to a lot of information, both legal information because they're caring for people and as well as their own employment rights, their own processes, their own normal everyday kind of working processes that they would expect as part of their employer kind of relationship. So you're looking at how AI can support that availability of access to information, knowing that some of those people may not be as digital as you are, maybe not have access to digital technology in quite the same way as we both have, or may not speak English as first language, or want all of their service in Welsh language as first language. You start to add more and more layers of complexity, which is what makes your role more and more complicated, doesn't it sound simple, but actually it's the diversity of what you're dealing with is really complex. Absolutely, and you know, we're speaking short of that in the services kind of inclusive, in all kind of meanings of it as you mentioned language as well. So yeah, it's quite a lot of discovery work to really understand what we could deliver and what would be acceptable really for people accessing the service. Okay, so can you take us through as a state, you just talked about discovery there, but maybe you just take us through the main stages of how you started the project and what are the kind of challenges where you came across each stage, and conscious some people won't necessarily know the process, so you could just kind of give us a little synopsis of each area I think would be really useful. Yeah, sure. So I worked through the project through four stages, so discovery, where we would be doing research, some of that would be just desk research to really understand the problem, make sure we're looking at, I guess, the problem in the right way, and some of that might be interviews and speaking to the members of staff who work within the service to understand their behaviours and how they might access that information already. Then the second phase I looked at defining the problem and really really looking closely really at making sure that the scope was right and that what we were going to be able to deliver would be realistic and feasible. So within that, you know, we did some, I guess, now looking at our technical options in terms of I guess cost as well, we're a public service, so that's a quite an important aspect, looking at all the possible ways that we could deliver it, and working with the teams really to look at all the different ways we could do it. It may have turned out that actually using AI or something really techy wouldn't be the right approach. Then delivery, really kind of pulling people together to like move this project into a build, deliver minimum viable product, so you know, what is the smallest thing that we could do to solve the problem and be able to test that? What is the problem? What was that in that stage? That stage, it was about, I guess, what we found within, I guess, the delivery of the instructions is that people do use lots of different ways to sort of ask a question of senior team members to try and solve a problem when we have actually a body of information available that people could access, so what we're trying to do is free up capacity within the teams and really help people to self serve information themselves, which is quite an important point, and I think AI, the use of AI is in the use of lots with people and not always for the right reasons, so people are quite nervous, so I think in terms of solving the problem, we found that we kept, and not just look at the technology, we had to look at how people would react to it, how we could really sell the benefits, really, are you doing something in a different way, which has been feed into the later stage of the project around implementation, getting people to accept the changes, really help them take ownership of this new way of working. That's great, and that's a different kind of way, isn't it? And ultimately, that freezer staff that would normally be responding to those inquiries can now concentrate on doing other things, which ideally will be working with wider community with the caring community, so they're not just dealing with quite a laborious, same question, same answer every day and every week. Absolutely. That's all kind of automated, okay? So in terms of, and those questions coming through can be anything from payroll, holiday leave, general inquiries around registration details, and such like. I'd say some things around the registration detail, a lot of questions about qualifications, because there are lots of qualifications available that somebody could use to work in the sector or they sound like they should be able to use to work in the sector. So that's like a whole of the thing, really, because there's a lot of information available. So then developing the AI, what does the AI do? People have got different interpretations for AI is and positively, negatively at the moment. Yeah. So what we were looking to do was really make the most of, I guess, existing large language models like ChatGPT and our specialist knowledge so that we hold internally. And what we wanted to do because we know that large language models are really good at generating information, but kind of left with its own devices, it will give you the emissions really to give a user an answer. So sometimes that might mean if they don't really have the right answer, they'll just tell you something. So it seems like it's right. But what we wanted to do is to use that technology, but then I guess use our information about the work instructions to, it's kind of the main brain for that information. So we would use the ChatGPT and every time someone asked a question for it, it would check our body of knowledge before generating the answer. So your company policy or your most government strategy or whatever it's aligned to? So there's always a reference point. Yeah. So it's kind of the main reference point. And I think that way we kind of ensure that it's always using the correct information, it's kind of closed off. We're able to keep that up to date without having to build a whole language model ourselves, which would be really expensive and complicated. We're able to just use that and then build on it really to make it super accurate for our needs. And those are the things really, I think especially when you're working the public sector, these are the sorts of assurances that we need to give our leaders really. And actually we can make things safe. We can make sure that it's not going to give any answer. These are the things that we're telling it. We have that control. And actually when it really doesn't know something, we tell it to tell us it doesn't know. So again, that's safe. It's better that it says I don't know them gives you any answer really. Yeah, yeah, that's great. And the confidence that you've subsequently given the senior teams within health and social care, has that been one of the barriers? Is that has that been one of the challenges or is that I definitely think within our organisation the leadership has been really positive and really been driving for us to kind of look at new ways to use technology and to innovate. But the kind of on this journey with us, we're trying to share them actually. We can do these things and this is how we think we could do it. For our capacity now and the kind of problems that we're kind of looking to solve. So I think from that sense, it's been really great. I think there are some other areas where we've had to really kind of really sell the way that we want to work and how we want to move forward because we know that the amount of work isn't getting smaller. We have ways that we can kind of streamline some of that. So it's kind of more kind of a changing hearts of minds that we've been going through. But I definitely say that in terms of the leadership, they've been very willing, very open and to be like a real driver for us trying to share this way of working without the workforce. It's really interesting that you've changed, you know, you started this conversation really about digital and AI. What you're talking about in terms of the impact of what you're doing and the things that you've learned on that journey are human things, aren't they? That need for leadership and the need for kind of confidence. And you're talking about we quite a lot now, which demonstrates the team ethic that is behind the whole project. Yeah, we're definitely, I'd say, any of the kind of challenges that have come across so far within the project have been about people and really just setting up the right foundations for it to succeed within the organisation. So people know the direction that we're heading in. So for example, that's developing an AI strategy. So people know, have a really clear understanding around what we're thinking about when we're saying it. We're not talking about introducing robots and things like that. We're talking about thinking smarter and we're setting the parameters around how we're willing to use those technologies. I think that's really important. So we're all working from the same direction. And I think it was around skills and helping people feel fully competent when they're kind of using these sorts of things and how they might approach it. And to ask questions to help feed into where really important. And then just giving people the confidence and to know that they can actually, you know, we've built a really good sense of psychological safety and we need their feedback for these things to succeed. So they're always coming back to us with the right information. And a lot of that, you know, it's not quick or easy. And especially with the landscape changing so rapidly. But I think we're heading in the right direction really to make people comfortable. So, you know, we're heading in that way. That's brilliant. Where is the project now? Have you rolled it out? Do you using it things now or? No, we have built a prototype which works really well in English and Welsh. So we're really satisfied with that. We've got some other kind of like foundational work happening. I would say we'll probably see something in the next by the summertime really going into live. And then just working with the team to make sure it's 100% working in a way that they want. What about the impact that you're starting to see as you've gone through the testing stage? Do you see your users starting to engage with it or what feedback have you had so far? Yeah, they have actually engaged quite well with it. I think what we have found is that people, the way that we use this conversational, I guess, AI actually suits people a lot more than having to fill in a form or to do something else, which was important to us actually because they were already using this conversational style. So we didn't want to sort of introduce something that would be totally different to what they were used to. It's kind of matching where people are now. It's able to make people feel a bit more relaxed in using it as well. And I think that's important again, just creating the right environment. And people have given feedback so far because they want, they actually wanted to work and for it to make things better. So that's been really positive as well. That's a main. And I guess the kind of people that are working in the sector by the nature of who they are, they integrate, they're working with people, they're speaking to their clients all day and that's the way that that's their skill set almost, isn't it? So it seems to contradict to then go through a kind of bureaucratic or digital kind of process, seems almost a kind of contradiction of who they are as a group of people and the values that they bring to the organisation. And then in terms of feedback from your own organisation, you said you've got buy-in from your senior team and that seems to be going well. What about the impact on yourself? What have you learned about this whole process? Yeah, I guess you've mentioned it really that a lot of these things are about, I mean, work with technology itself is actually much more straightforward than I think people have you believe. It's really working with people so that they can feel a part of the journey and to take ownership, make sure we put the right things in place so they're able to be successful within that space as well. So lots of the challenges are definitely about working with people, building relationships and I guess selling things that show them the benefits of how it can make an impact to their day-to-day working life. I think that's really great. It's a service design, half of it is about facilitating things or being a steward for change so it kind of fits nicely within the day-to-day job. Yeah, yeah, but they're all human skills aren't they? They're all, they're all it can implicit human characteristics. Sharmayne, it's been amazing, thank you. There's really been really useful. Thanks so much for your time and the success of the project. Thank you. If you'd like to hear any more of the Change Management podcast please subscribe. Thanks for listening and thanks again to our guest, Sharmayne. To find out more about this podcast and other business services at usw, please visit southwills.ac.uk/business.