Conversational AI and how to implement it successfully

Louise
In today's episode, I'm delighted to host Sarah Topping, senior Intelligent Automation Consultant at Ciclum, where she works on creating state of the art, user centric designs for conversational AI. Today we dive into the best practices of designing an intelligent conversation agent from the ideation phase to the delivery. We talk about what are the requirements, what are hurdles, and especially what not to do in order to avoid churn due to a bad chatbot experience, which we can all probably relate to. Sarah also shares how conversational AI is interlinked with other automation technologies such as RPA, and how their full potential can be unlocked when choosing the right technology for solving the right problem. Finally, we talk about her thoughts on having designed a learning course in the Bots & People Academy where Sarah provides a great hands-on learning experience all about conversational AI. Now, I hope you enjoy this episode.

Louise
Hello and a really warm welcome to you, Sarah. I'm really happy to have you on our podcast today. Please let's start by you briefly introducing yourself and tell us what it is that you do.

Sarah
Yeah, thank you for having me. Firstly. So my name is Sarah Topping. I'm an intelligent automation consultant at Ciklum and I work in Conversational AI. So there are two parts to my role. I work closely with clients to help them identify customer and employee interaction journeys and translate them into really great conversational AI journeys. And I help them define best practices around virtual agents in terms of implementing user experience, conversational design and also help improve their existing automated customer journeys. And then my second part of my role is very much more in Ciklum. So I'm the Conversational AI leader as I mentioned but also the contact center lead. So I'm involved in growth and development of Conversational AI at Ciklum, so defining the strategy, developing collateral methodologies, leading partnership with vendors and growing the team and then also supporting our go-to-market teams with presentations for clients for any sort of new potential clients that are coming in as well as well as existing clients as well to help extend their sort of repertoire of technologies as it were.

Louise
All right, I see you have a very broad scope of things that you do at Ciklum, but what I'm really interested of course is especially the part where you're talking about conversational AI and how to develop that for your clients and we'll touch upon that later. But before we do that, maybe let's get a little bit back on how you got your career into this point. What's your educational background and what got you interested in this very specific field in automation and conversational AI?

Sarah
Yeah, definitely. So if you'd have asked me 20 years ago or even ten years ago, I probably wouldn't even really know what conversational AI is in terms of my educational background. I did a very science based background. So at college I studied psychology, biology and environmental science before then going on to university to study zoology. One of the key things I always found in that was my interest in both sort of human and animal behavior. So I always went down the psychology. It kind of helps with conversational acts. It's really about thinking about people and how they react in different situations. I think that is like a little bit of the educational background that comes into play with it. But generally I left university, I wasn't really sure I wanted, to be honest. I started working at HSBC. We have an assistant role and it just so happened to be in the digital team. So I was working for senior leadership role as their assistant in the digital team and that's where my initial spark was sort of ignited. So I'd worked a lot in customer service roles growing up in shops and restaurants, but I'd never really been exposed to how companies communicate with customers digitally.

Sarah
So there I worked a few different roles at HSBC before moving into a new team that was formed about five or six years ago now, and it was a conversational banking team and then specifically into the chat bot team. It was very much a steep learning curve for me and one that I sometimes didn't really feel qualified was. What do I know about Chat bots? But it was more about that user experience and learning the technology side was great. I happen to do quite well at it. I'm hoping to have a say. I'm a developer, but I understand enough to be able to take the technical principles and translate them into business principles and I always like to think of the user. So I thought, you know what, actually I really enjoyed this. I worked really hard. That seemed really great results from the work we were doing, like customer service scores were really great and so I thought, I really enjoyed this. So I just really made that sort of my passion and really, like I say ten years ago, I wouldn't have known what AI was, but now I'm very much an advocate for it and if anyone asks me about it, I can talk for it for hours.

Sarah
It's really great and it's been probably not maybe a typical career into some of the automation space, but it's very much been one that I've enjoyed.

Louise
Nice. Yeah, that sounds brilliant, honestly. And as you said, you can talk about it for hours. That's exactly what we're looking for, to hear you talk about that more. But I think it's great to see how also from a very different background, you can move into the space and actually learn about it. And yeah, I would like to know maybe also now that you're currently at Ciklum, what makes it special to work there? What do you love about it now?

Sarah
Definitely so working at Ciklum, I gotta say it sounds a bit cheesy, but first and foremost, it's the people. Both at Ciklum but also all the clients I get to work with, I get great exposure to different working practices, different companies, different space. Secondly, the trust that the leadership team have as me and my team members, they know that we know our areas, so they really put that onus, on us. You know, it you tell me what we should do and we'll support that. So that trust is really important. And then I suppose the innovation and exposure to technologies as well. And how about they worked hand in hand with Conversational AI? So before working at Sickler, I was probably a little bit siloed into conversational AI. I was like 100% focused on it. But now I've learned so much more from my peers and just learnings that I've done on my own around RPA and process mining workflow technologies and I'm now seeing how like how that technology really helped Conversationall AI just not be a standalone tech, actually how you can add all these great features in to really make like a robust and sort of hyper automation.

Sarah
It's really taking it to that next step. So that's what I've really enjoyed about working at Ciklum. It's really helped me develop, but it's also then you know, it's helping all the technology stacks work together and really sort of harnessing each area's skill set to sort of make each area then not just like on its own, but yeah, it's really great for that sort of side of things.

Louise
Right, OK, I think you're touching upon an important thing that I wanted to ask you anyway, so maybe also the listeners that maybe don't know Ciklum, maybe you can explain a little bit what you do as a company. What does the core business and where is Conversational AI anchored in that? And also how does it maybe then relate to the rest of the automation space? You've already mentioned that you've kind of started getting into the other technologies and see them all sort of how they work interlinked, but maybe just break it down for us a little bit. What do you actually do at Ciklum?

Sarah
Yeah, definitely. So Ciklum, it's a global organization. We have very strong design and delivery credentials. We have digital engineering as a core pillar, as our brand and DNA. But then as a company, we are also committed to building solutions that are resilient and sustainable. We don't want you to move away and then it falls down. We ensure delivery excellence and value for our clients through our delivery centers across 20 locations globally, which is great, continuing to expand. And then we've got three capability areas that sort of stand across our consulting arm. So we have the first one which is sort of the create custom products for our clients, so that's sort of the product team. We then build data and analytics capabilities and solutions. And then thirdly, we have a really strong intelligent automation capability and that's where Conversational AI sits. The AI department is focused around working with clients to sort of increase their efficiencies, enable innovation. And along with conversational AI, as I've touched on, we have experts in RPA, IDP, so workflow low code and no code and then also process mining. And we make sure all our consultants, such as myself, we all have generally specialisms.

Sarah
We are trained across the multiple technology as well. So you're not sort of just blinked into that one bit of tech. We can really go into a business area then and challenge and ensure that the best solution is proposed. And this is where we often see some really exciting and effective solutions. So we start using RPA and IDP to really supercharge the value of conversational AI and we utilize all the different tools. So yeah, that's what Ciklum is and how we work in a bit of a nutshell.

Louise
Right, okay, great. I think it's really important, as you said, to also look at the problem or the thing that you want to solve and then decide which technology to use. And by all of you understanding them and working together, I think that's a great proposal. So what I really want to now do is dive deeper into the topic though that you're a special expert in, which is conversational AI and kind of let you unpack a little bit the process behind that. So you are kind of creating it from the idea, OK, let's have a chatbot, let's say, and then you deliver it until it's actually up and running. So yeah, maybe you can just talk us through that a little bit.

Sarah
Definitely. So for those that don't know, conversational AI is a way of offering conversational experience. It imitates sort of a conversation with real people through AI algorithms and you put them in places that are really convenient for people to ask a question and have tasks completed. So these could be on websites, they could be on various social media channels or in an app. And most people have used conversational AI in their day to day life. One of the problems that I find is that most people have used it, but most people have had a bad experience with it. The technology has really advanced in recent years, but there's a lot of companies that still have quite the oldest style tech that's not really changed in the last few years. So there's a lot of unfortunately badly designed and implemented virtual agents out there. So like yesterday I used one and obviously going with a bit of a lens in there because it's my day to day life. But I used one yesterday to cancel my dog's insurance. It was not very new. The bot wasn't able to deal with it and even though I can bet that's a really common use case for their customer service department, it said it couldn't deal with it, it had to pass it to a human.

Sarah
It then told me it couldn't transfer me to a human agent as they're opening hours with 09:00 a.m. To 05:00 p.m.. However, it was 09:45 a.m. So it not only failed to do the action for me then, didn't know the times or they changed their opening hours but not updated their virtual agent to let the customer know. So it's cases like that where actually that's a really great use case for conversation. The possibilities are out there. Could have authenticated me as a user connected to back end system, updated my account and canceled my renewal in a few minutes and there will be absolutely no human interaction. However, I think what the company was trying to do, and this is quite common unfortunately. We got to get out of this mindset when it comes to conversation AI. It's because I imagine the company wanted to retain me as a customer. So that's why they want me to speak to a human agent and not take action through the bot. But they've not designed their journey well. And then all I had done was waste my time and annoying me as a customer. So my mission of the company has plummeted.

Sarah
I probably would have canceled regardless. The price wasn't great but I might have returned in the future as a customer in a year or two's time when my next one was up for renewal. But now all I think is well actually am I going to go to that company because I've had a bad experience with them. So actually that customer satisfaction has actually made me not going back to that company for probably several years now. That's why the user experience is absolutely key to conversation and that's what I'm passionate about. We don't want people having those experiences because it's not a good example of it. So when I'm looking then at a journey, a company once comes to me, they want a chat bot. What do they want? It's thinking about these requirements ahead of it. So does it need to be 24/7? Will there be a human agent support? What channels are you going to be on? Think of the really great happy customer path but then also think of the really unhappy customer path like what can make it worth and plan for it and then what are your success criteria? And then key to that point there, does it need to be authenticated if you're doing a customer journey like that?

Sarah
Can you connect to a back end system? Think of all these things up ahead so that actually you can have a really cool end to the enda sort of user journey and then for me it's then moving on to thinking about the prioritization of what journeys you're going to do. So you've got all these requirements. You understand what you need to get to. But what actually you're going to do. What's your book going to do. What are your common queries that your user is going to ask what can be easily answered without a human being involved. Or actually if a human needs to be involved. Can the bot sort of do some prerequirements. Gather some information to make the agent's journey a little bit easier on the agent side and really understand what the cost benefits of doing the project is? All businesses are business at the end of the day. They want to improve their customer journey, but they also want a return investment. They're going to spend all this money, what are they going to achieve? So really understanding that so that you can set a benchmark of what you're trying to achieve.

Sarah
The next step for me would be making sure your resources are aligned. It takes various different parties to get this together, from a project manager to an engineer to a conversational designer. Have you got all those people around to get this done on time? And then actually, then you get to sort of building and training your virtual agent. And this can involve like, creating content. It's a conversation design. It's gathering data for the NLU so you can train the brain of the virtual agent, essentially. And it's on the testing period. You can't skim on that test, retest. Does it work? Why isn't it working? Train it again. And that can take a bit of time. Once you've got through all the testing, which can be sometimes like it seems like it's taking forever, but actually it's really important. It's like you go live. So great, when are you going to go live? Like really plan when you're going to go live as well. Is there any other events happening in your business? You don't want to put a bot live, especially if it's a very new bot on, say, an e commerce site on Black Friday, because you know the demand is going to be insane and you might not be able to cope with it.

Sarah
So try and pick like a quiet period where nothing else is really happening and then you have a really sort of unbiased opinion of how your bots performing because there's no other try to limit your variables and then after that it's the case. To that point is it's monitoring its performance? How is it doing for some next steps? And this is vital a lot of companies don't really think about is that continual optimization, like reinvesting in product, you might have saved all this capability in terms of human effort required. Take some of that and reinvest it to keep it fresh, keep it new. What's working well? What's not working well? People not enjoying it. Is the language not correct? So they're really the key points for me to really understand. How are you going to do it? How you're going to execute this?

Advertisement
Why buy process automation at a high price when it can be built internally and sustainably? Surveys show that a lack of knowledge and skills is the number one cause of death of all automation initiatives. More and more companies realize that automation success is not only about expensive tool licenses, but the perfect combination of Bots & People. Therefore, the Bots & People Automation Academy offers over 50 learning modules to train employees and managers in topics such as automation strategy, citizen development tools and technologies and many, many more. Visit www.botsandpeople.com or click on the link in the description to read all about it.

Louise
Okay, so thank you first of all for unpacking that so beautifully. I think it's really interesting, first of all the example that you gave of a really frustrating experience and then seeing all the best practices that you suggest and how you can actually make it into a great tool for companies and use conversational AI in a really beneficial way. Maybe as a next point we can talk about what kind of use cases and projects you do. So in which kind of areas do you have these kind of conversational AI or these agents that you build and what are your sort of no-brainers? Let's say you've been talking about customer service interactions but also maybe some more complex cases now that it's not an advancing technology definitely.

Sarah
So as a general rule of thumb, anytime a conversation takes place, that's an opportunity for conversational AI to be implemented. So as we touch on customer service, first and foremost one of the biggest areas. But then we have internal company ones that we might look at as well. So employee help, so It or HR and then also e commerce, these are all the no brainers for me. They can offer some great value and to be honest, they can all be delivered really easily or they can all be complex. And the complexity usually comes more from integration with back end systems. So as one use case. So for example, if a company has got really old back end systems that can add a layer of complexity, it might add some extra time onto your journey. Similarly, if you have a knowledge base existing that you want to utilize, what a lot of companies find is actually it's been designed from an agent view and it's not actually suitable to be like sort of user facing. So they have to do some changes there to make it suitable to make sure the language is like no jargon or that sort of thing and language that would only be understood internally.

Sarah
Some of the other complexities around actually not being operationally ready. A lot of companies might have a voice channel or something like that or work on email and then they're moving to a quicker environment or from a voice to a text environment for their agents. So actually you got to train those agents up. It's a different way of working. They might be working with the same customers but it's very different. It's a big change for them and you've got to bring those agents on your journey with you right then also having robust data. So being able to train your natural language understanding, which is the brains of virtual agent. You might not have much real life data, especially if you're new in the journey of conversation AI or you don't even have like email or messaging as a channel. You have a voice and then you might have to do some voice to text translations and it all adds a layer of complexity and how people talk on different channels. It really changes and these are all complexities that can be overcome and they should never be considered a blocker. But just having an awareness of these things and making sure that you add time onto the delivery.

Sarah
If you're dealing with some of these challenges. It just really helps set expectations and avoid disappointment. So they're like the complexities. I come over quite often but there's none I would say that I've really come across where we've not been able to overcome it. It might have just taken a little bit more time. It's just really been clear with that upfront and especially if you've got senior stakeholders that you need to keep engaged in the journey. Just honesty sometimes with these things really helps because you're showing actually things aren't perfect, but we're going to get there and we're going to try and make them as perfect as possible.

Louise
Right? I think it's great that you're mentioning that kind of transparency and also managing those expectations in terms of the delivery. But I think what's an important point is that you're saying basically in any situation you will be able, if you have the right requirements and you set them up in the proper way, you can actually build this kind of thing for a lot of customers and a lot of use cases. So my question is I still face a lot of situations where I have to wait in a hotline for forever basically and then don't get really satisfactory way of someone dealing with my problem. And I feel like I could have just typed in my customer number somewhere and they could have managed it. I feel like a lot of times we're not there yet with companies making use of this and what would you say is maybe also holding them back? What do you think are the biggest challenges for companies with regards to conversational AI and maybe also to add a layer to that? How do you think the covert pandemic has changed that?

Sarah
Definitely. So the main challenges I would say are those back end systems and just really using some other technology if you can't integrate into them straight away to plug the gap. So can RPA be used or can low code platform such as an appian technology be used to plug the gap where two technologies won't talk to each other? And that's where you will see like when you say putting your customer number in, you should be able to do that. The technology is there and we are in that state of actually the technology exists, we can do it, but actually sometimes it's a bit of a slow process to actually get it rolled out. It's like prioritizing your backlog essentially. Like when it comes to what queries you're getting in is that query that requires a back end sort of development or another technology to fill the gap actually is that one that is really up there in terms of the volume coming through so it's worth the value. If it isn't, it then comes back to that conversational design piece. So don't try and keep people in a journey if you know you can't support it.

Sarah
Get them to a human agent that's what you know human agents are there for. Don't hold people in a virtual agent if you know you can't deal with it. So just really make sure that design is there and get to that point that actually quite honest with where your technology is and get to that point. Your virtual agent is still learning, it's training it's there as an assistant, but if it can't help, let's get you to a human where possible. And it is really just on that journey of companies taking that sort of crawl walk run and then my next stage of optimize should be on there as well as a continual crawl stage. Just keep going, build it up and have a planning place. What do you want to achieve? So yeah, another one would be a challenge that is often faced. So you might have a project team that has really clear ambitions but then they can be internal red tape. So challenging any sort of internal governance. Challenge the status quo from your project team. Is it there? Actually it's needed, it's governance, you're keeping your customers secure or actually is it just there because that's the way we've already done something.

Sarah
Does it actually need to be there? So yeah, having that sometimes to challenge that and make sure you understand why it's there and actually it's a requirement. So actually it might slow down your development but it's absolutely needed to keep your customers secure or is it actually let's get rid of it. It was from ten years ago and it doesn't is really relevant anymore.

Louise
Right.

Sarah
The COVID pandemic absolutely shot conversational AI up. It was already in merchant tech and there was already a lot of companies implementing it, a lot of larger companies especially, but the pace of which then people were at work or working at home, it's very restricted. So you might have agents that they're online and then the internet goes down, especially there in some countries that might maybe don't have reliable internet connection. So actually you need something else to help support your customers. And it was great because it's certainly peek conversational AI lot of companies who maybe didn't know about it before were now aware of it, but a lot probably also implemented it at that time period in COVID because it was a necessity, it was like a kneejerk reaction, but they might have left it in place now. And now as we start to get back to our new normal life, is it still working as effectively? And that's why I say optimizing is important. It's go back and check if you implemented it in COVID, is it still relevant now? Is it giving you the results you want? And how can you keep working on it to keep improving it, to make it fit for now and in the future and not just what state we were in two years ago.

Sarah
I think it was a very key point in the conversational AI life cycle. It's one that when I mentioned some of those bad bots out there that sometimes they're like relics of that of two years ago that need to be sort of looked at and reinvested in just to make sure they're still giving that great experience.

Louise
Right, okay. So when you say that the topic started peaking and it really developed in that period, that also means probably that companies were looking to build up the competencies to deal with these things. And you also said it's important to reinvest your resources into maintaining it, into improving it. But maybe that's also something that is new to the organization. So I'm curious also to see what kind of requirements or what kind of competencies are necessary in an organization now that is trying to keep or to build virtual agents that is trying to maintain them. That also has this understanding of designing a journey in a good way. What kind of skills. What you say are new and that are key to make this work in a successful way.

Sarah
Yeah, definitely. So I think it's putting the user experience first and foremost, put yourself in the user's shoes. What was done maybe for an organizational benefit in COVID, again, is that still relevant? Making sure you're not pushing customers down a route that suits you organizationally, but actually might cause massively customer dissatisfaction. Customers, when they're talking to a company, they don't care what department they have to go to, they just want their problem solved. So if you like, oh, well, this department has a chat bot that talks very differently to this one, or it has a different experience, it can get confusing. So making sure you have you know you thinking of the whole journey, not just like a siloed organizational approach. And again, it is that reinvesting of time. And if you did have something already. You've got so much data to hand. Like really utilize it. You know. Really look at it and say. Right. I got all these customer assurances that are not matching. Or. You know. The matching the wrong intent on a virtual agent or a chat bot. Actually use it and take it away and analyze it and use the other tech that I've touched on.

Sarah
You know, there's process mining that can really take some of this data and take some of that away to make sure that actually are you doing the best for your customers from all these data that you've got? Actually, you've got 100 people going on a journey, but 100% of them are falling out at a certain step. There's a problem there. Like, what is it? Have a look at it. Is it that it's not fit for purpose? Is the language not right? So it's taking that to the next level and then it's making sure that if you've got maybe a more FAQ style bot, like just question and answer, it's not very conversational. Can you take it to the next level and make it more conversational? Can you integrate it to a back end system that before you maybe didn't have a need? You can see customers are using it. Can you connect it to your CRM system? Can you connect it to a payment system? It's all in one channel. And you're not deflecting customers from one channel to another to another to do something. Just do it all in one place if you can.

Sarah
It's really the key to the next steps, is using other technologies to look at the results and make them better, but then also looking at integrations to other systems to give that end to end journey rather than just like, one piece of it.

Louise
All right, thank you for that. There's another topic that I would like to talk to you about because it's kind of something where we also cooperated with Boston people. And you recently published a course in our academy on the topic of conversation and AI. And I'd really like to hear a little bit about your experience with that and also what you kind of really try to focus on in that and how you think it's best to learn about this topic. How can you enable others to follow all those best practices that you've mentioned before. So, yeah, just tell us a little bit about that experience as a trainer in that topic.

Sarah
Definitely. So it's a very exciting journey for me. It's not something I've done before, but we were introduced by a colleague, I work at Ciklum. So, yeah, it's great. I really enjoyed it. Great team to work with. So that was great as well. And it was a very easy journey to create the content because it's close to my heart. What I really enjoy, and that's what I wanted to get across on the course, is actually, I go on about it all the time. It's a user experience because I always say you can build like, the best brain of a bot in terms of an NL unit, and it's, don't get me wrong, vitally important to have a really intelligent, natural language understanding your bot. But actually, if your design is not great because you've not considered your user experience, then they're never going to get to that point because they've fallen out the journey. Because they're like, well, it doesn't make sense, I can't read it. Or it's full of jargon from a language perspective or not taken into account accessibility challenges some people might have. So the national reading age in the UK, I think it's about ten or eleven year old.

Sarah
So making sure that you're really targeting everyone and not leaving people behind on this journey, making sure that you account for people that might have need screen readers so it's accessible for all, the chatbot has a timing in it so that screen readers can keep up with the conversation as well. And then the really key and important things for you I like to make sure that are always incorporated into any sort of project that I work on it's making sure that it's a great user experience and that it's accessible to all. And that from your color design of your engagement window, certain people with different challenges might not be able to see certain colors. So really look at what all aspects of the design and the user experience to ensure accessibility. And then I really was interested in. You know I came to conversation, like I mentioned, not probably a traditional route. So if you find an area that you're really excited about and that you want to learn more about, that's why doing a training course like the one that I've done, I would really say invest that time in yourself and your development, push yourself outside your comfort zone and do something new.

Sarah
If you're in work, maybe ask a manager or employer if it's suitable that you take some extra responsibility or maybe an access to a training course. Because a lot of the great people that work in conversation either, I've learned over time, actually come more from a customer service background, especially like so they've got that wealth of experience, they've worked on the front end with the customer or the user. So actually they bring a lot of talent, but they might need upscaling in other areas such as development or, you know, actually putting their all their knowledge down into a workflow. So yeah, doing pieces like that is why I really enjoyed doing the Bots & People collaboration in terms of the course on your academy, right?

Louise
Yeah, I'm really happy to hear that and I think you're touching upon many different things that are important. But one is that really enabling people also that have an existing skill set and taking it to the next level to go with also the digital transformation to go with the, with the age that is currently happening and using those skills in a new way. And I think that's a great thing that we're always also trying to preach, as we said, looking at what is the customer need, you only know that when you really are able to think like them or to put yourself in the shoes. And I think it's similar when we talk about other automation projects. You're trying to automate a process, you want to make sure why is this process existing? Does it work that way because we've always done it like that or because it has an actual purpose. So making sure that you are actually creating value with this digital thing that you are creating, does it have a purpose? And doing so by being enabled and having all those skills I think is a great way forward. So I'm really happy that you shared all this knowledge and I know there's a wealth of things that you could talk about and unfortunately we have to look at the time, everything that you shared so far.

Louise
Of course there's the course to look into and many resources available. But before we close this episode, I'd like to have some quick fire questions from our question vending machine. And you can just answer what's top of mind when you hear the questions. So the first one, maybe you've already answered this one. What is the biggest fail you've ever experienced with a chat bot?

Sarah
Yeah, in my personal experience, it is one like that where you just like why you had such an opportunity and you've failed to do it. So yeah. That's probably for me. One of the big ones is when there's an opportunity and if a company is not taking it or for whatever reason and that's it because it's just annoying and it's annoying you as a user and then that puts a bad if that customer then just goes to a human agent. That customer's already annoyed and then the agent has to deal with it. So then you've got the agent experience as well side of things, which is vitally important. So why are you annoying customers to then upset one of your team? It's not necessary. Right?

Louise
Yeah, I think we can all relate to that and I think there's nothing worse than having a bad experience like that. So yeah, definitely understandable. Another thing that you touched upon slightly is what drives you in your work.

Sarah
Yeah, definitely. So when you've seen a project and you're in it day in, day out and you're so sort of focused on it, you put so much effort and time into it and sometimes you might even thought this is never going to come to light. Like there's problem after problem but then actually see it go live in production and it's a success. That's the bit that is that moment when you're like oh my god, people are using it and it's working and they're using it well, they're happy using it and it's that piece like all the time and effort and maybe even times like when you are stressed, it's worth it for that moment. That's really what drives me.

Louise
Great. Okay, and then I have one last question. What is your favorite activity to create a positive culture at work? Because you said you really enjoy it to work where you are and have a great team, but what is something that you like to do and maybe something that you can share with the audience that you recommend doing?

Sarah
Definitely. For me, it's all about communication and then collaboration as well. So communication doesn't have to be sort of work focus, it can be fun activities, it can be going out, it can be virtually or face to face, but then get to collaborate on projects as well with team members. Like that's really great. Especially if it's an area that you're interested in or maybe you've not really had much exposure to it. You're continually learning. So it's that as well. That's really great.

Louise
Right, okay, well, thank you so much, sir. I think I learned a lot today and I'm really curious for more. Thank you so much for sharing everything and yeah, for coming on today.

Sarah
Lovely. Thank you for having me. It's been great. Thank you so much.

Louise
Thank you.

Advertisement
Thank you very much for your time. We hope you enjoyed today's episode and that you were able to take away a lot of interesting information and insights from our guests. For even more insider knowledge freshly coming from the core of the automation industry, feel free to listen to our other podcast format called Automation Insider. This is hosted by Nico Bitzer and Andreas Zehent. Or if you prefer to read, subscribe to our biweekly automation magazine.

‍
If this has intrigued you to learn more in the area of process automation and this is also interesting for your company, then feel free to visit our Automation Academy and get advice on Upskilling. You can find all the necessary links in the description. See you next time. Your Bots & People team.

‍