#108 Next-Level Business...AI & Conversational Business Applications

Hello and welcome back to another episode of the Process and Automation Podcast with the Automation Guys. Yeah, as always with me here is Arno. Hello, Arno.

Hey, Sascha, how are you doing?

I'm good, good. Thank you very much. So yeah, we are back. With another episode. And yeah, we have just passed the half-year point of 2024, and yeah, with that, we, we looked back a bit and reviewed a couple of the topics we had covered, and we thought it would be great to balance it out a bit again. Um, yeah, we haven't spoken about conversational AI a lot recently, although AI is everywhere. So, um, yeah, so with that, we, we are doing that today. So yeah, many of us are very familiar with chatbots on websites. So this little pop-up thing there at the right bottom corner, which asks you a lot of questions when you join that website. So, but there is more to chatbots these days, especially since the huge advances we have seen with AI and GenAI especially. So yeah, that means we are not really talking about chatbots anymore. We talk about conversational AI. And also going further, was that term would be conversational business applications. So it's quite a big thing if you think about you could have a full business application which be fully sort of covered by conversation only. So that's quite nice to think about. And yeah, this is something we talk about today. We will talk about how we can introduce Conversational AI to employees, to the workforce, and how that helps them to streamline tasks And of course, that's one of the big topics as always here, enhance productivity and empower employees. So I guess, Arno, should we just jump in with the first point?

Basically, sort of, yes, thank you, Sascha, sort of the employee perspective. Yes, exactly. And I think, you know, virtual assistants and then sort of embracing virtual assistants is becoming a bit more sort of mainstream now, and, you know, kind of the change can be daunting. You know, we look at sort of virtual assistants and it's AI-driven and people think, well, is this virtual assistant gonna displace me? You know, so I think the reality is that conversational AI has to, you know, has this potential to really enrich the role of the employee rather than replacing it. And simplify daily tasks, for example. So we might have stuff like, you know, customer-facing virtual assistants. So these are sort of intelligent virtual assistants that could manage, you know, initial routine customer inquiries. And, you know, that frees up the human agents to handle far more complicated issues that require sort of a human touch, human empathy. And creativity. And again, this shift allows employees to build, you know, stronger relationships with customers while the AI sort of handles the repetitive task, you know, the kind of FAQ, you know, what's the outstanding balance on my bill, for example, where's this product, and I returned it. So there's a lot of those sort of, I guess, those transactional use cases that a virtual assistant can field for us. And then if we do have an escalation and we need sort of a human touch, then of course we could transfer that, effectively, that chat to a human, and that person can then talk to the human and that sort of, like I said, you know, use creativity and empathy and then try and resolve, if it was an issue, it that way. And then of course, you know, just thinking about how conversational AI has got this potential to make people's role better and also to simplify tasks. The next one is sort of looking at internal AI systems. So these are not kind of customer-facing ones, and these are AI assistants that get deployed in departments like HR, IT, IT. And, um, you know, these types of internal AI assistants, um, handle sort of common queries from staff. Um, you know, where's the support ticket, or raising a support ticket, or what's my leave balance, you know, these types of scenarios. So, so that sort of frees up internal staff to focus more on strategic initiatives rather than fielding any sort of transactional questions. Could easily be fielded by an AI assistant, for example. So you could train these assistants to be very clever. You know, they can reach into the payroll system, they can reach into the sort of leave management system, expense management system. In the IT space, it could be things like, you know, I'm locked out of my account, can you reset my password? It just means that those types of scenarios doesn't have to involve human intervention, and it just means it scales as well. So if you've got a very large organization and you need scalability, you could have these virtual assistants at scale. So if you've got a very large workforce, things like password resets can then be handled by the AI bots without sort of, you know, applying human resources to it. you know, it's far better to deploy humans to understand why passwords are being reset rather than actually resetting the passwords. So that's, you know, that's kind of like the idea here is to free people up to do some sort of root cause analysis and use their creativity to understand what the problem is. And, you know, there has been a stat where they say that if you do deploy these conversational AI bots, you know, you could free up, you know, up to 70% of people's transactional time that they spend on these types of tasks, which is phenomenal. Again, if you've got a very large workforce and you do the numbers on that for these sort of transactional requests that typically comes through the door, that could be a significant saving for a company, for example, when they deploy this technology.

It is particularly when you think of companies are scaling, growing. So of course, it's not necessarily about— when we look at the 70-70% figure here, so we can save a lot of staff potentially, but now it's, it's very often the other, the other way around. Companies are scaling and they're looking at ways how they can scale in the best possible way. And yeah, and obviously it doesn't make— it doesn't really make sense to ask HR like 100 times a day the same question where it's like super easy either you find it yourself in your employee handbook, etc., or just chat. Yes. Which feels really, really human. And yeah, yeah. So I think that really makes sense to me, and I see it really going, going much bigger going forward. And yeah, for that, let's look at the practical applications of conversational AI. So we touched on obviously customer support, and that, that that comes to mind a lot when it comes to chatbots and conversational AI straight away. You're answering these common questions instantly. And yeah, this, it is even beyond that just normal inquiries these days. So I think insurance, it cannot just answer certain queries or what is my cover, what is my policy level, et cetera. It can actually go through the whole claim in one sort of conversation. So, and this is also customer support, but This is where claims management, for example, connects with customer support through one sort of good conversation. So yeah, it can be complex already, but as long as it follows, or maybe not too complex, it can be valuable, but it follows maybe a good set of simple rules. So yeah, it can really improve that efficiency in customer service. And yeah, I as a customer would be extremely happy if that would be on my insurance website to go through that one. So I wouldn't be bothered about, yeah, oh, I want to speak to this person to go through. No, so I think it's not required. Yeah, IT help desk, we mentioned that, password resets, but it also goes a bit deeper potentially because it can really go into the backend systems and can actually do things for you. So password reset is just a simple one, but it can pull out other information and move the ticket further, etc., etc. So HR assistance, we touched on, yeah, it handles sort of all the frequent HR queries. But yeah, what it then allows us to focus on, on sort of more talent development, think As well, client— not client onboarding, employee onboarding. There's lots of stuff to do where you want to be a bit more focused and really want to drill deeper into that, especially when it comes to like important roles you want to hire. You need that, that real human knowledge and focus. So, so if that frees the HR department while going through onboarding and recruitment, that's huge. It's a huge benefit. Yeah, so, but otherwise, obviously the assistant can guide all the new employees or prospective employees through that process. It can manage sort of the whole sort of offer letter generation and guide you through it.

Yeah.

So all in one seamless conversation, very powerful. Yeah, sales enablement is also a very practical thing. AI can help the sales team because it is connected to the customer insights, the AMP system, and it is also connected to market data, for example. So it really enhances that sales team's ability to get that extra information to maybe close a deal, win a specific pitch, or nurture the leads. But it's great to have that assistant on your side in form of conversation. And of course, employee training, it can enable your team to look at specific content and knowledge in a bit more personal way because you're starting communicating with that conversational AI, with the Conversational Business Application, and it will only serve the relevant topics which are relevant for you at that time, at your stage of the employment. So yeah, really, really powerful stuff. Yeah. Conversational AI absolutely goes beyond the normal sort of chatbot. Yeah. So, but let's look at concerns. Arno, there's always a concern.

Oh yeah. Yeah. So I guess it's always about kind of fostering that adoption and addressing the concerns that is this thing going to take my role? Is this thing going to replace me? Like I said earlier, I think that when you look at the use cases that the conversational AI bots actually fulfill, provide, it is, you know, it's always going to be, let's look at where the higher volume kind of transactional tasks are, let's give those to the bots. You know, if somebody does a password reset 100 times a day, it's quite a boring thing to do. A bot can do that, that could free me up to actually understand why am I resetting passwords. Are people putting the passwords in incorrectly? Is the policy wrong? So it's not kind of displacement, it's sort of just really enriching the roles. And I think that when you look at adoption, you have to actually position this as exactly that. So when you sort of integrate conversational AI, you have to address these common concerns. So job redundancies, for example, you know, it actually enhances productivity by automating these routine tasks, like I mentioned earlier. And it just frees people up to focus on more meaningful work, right? I'd rather sort of get to the root of, you know, why the password policy is wrong, rather than kind of resetting passwords. You know, that's not going to replace me. That is just going to free me up from a transactional perspective, if I don't have to deal with these tickets. To focus on that or to, you know, focus more on kind of their security strategy. Also just means that if the team sort of scales, if there's more people joining, so 200 or 100 possibly says from 200 in a year, you know, the conversational AI bot can actually feel that. So I think the whole kind of job redundancy needs to be debunked to say, listen, this thing's not here to take your job. This thing is here to make your job easier so you can focus better things, that's more meaningful. The other thing is also resistance to change. So obviously, what we need to be able to do, drive and foster adoption, is communicate the benefits. You know, we have to do that very early on in the program. And, you know, we need to provide support to help the employees to adapt to the change. But this is all the change management, you know, how's this thing going to work in my new day-to-day life? You know, where's the place for conversational AI in that? You know, what does that change management look like? And then obviously, you know, from, again, from another concern might be sort of the changing of roles. You know, if you look at sort of what type of support we need to provide, you know, the advice here is support career development and but encourage a culture of flexibility and adaptability. It is always going to be changing. If the conversational AI comes into the picture, let's be adaptable. Let's kind of foster that culture of, you know, we are going to evolve. We are going to kind of always adapt if there's new technology coming into the picture. And as long as people are content with it and it enriches the way they do their work. I think that, you know, that, that, that, that will drive that adoption. In terms of strategies to, to, to integrate conversational AI smoothly, you know, we, we talked about this in the past where we say, well, dip your toe in it, run a pilot project, you know, start with a limited rollout in specific areas and prove the value and look at this strategy, look what falls out of that and and just kind of keep on refining that. I think the other thing is also to provide good training and support for people. Um, you know, you give people practical training to, um, make AI approachable and, um, you know, relatable to, to what they do day to day. You know, don't see it, don't, don't, don't hide it away, sort of just put it in the forefront and make sure people really understand what it does, the benefit it can bring. Um, And then also from a strategy perspective, I think you need to define really clear objectives. You know, you need to say, well, we're going to bring conversational AI in, these are the target KPIs that we want to meet. We feel that this is actually going to make your life this much easier, so you will be able to focus more on things that add value. And, you know, we can constantly monitor that. And objectives could could well be that, I don't know, 90% of the password reset tickets we want to automate, and then 10% are the sort of edge cases that potentially we want to deploy engineer to actually help somebody because they are struggling and they don't really know why this password reset thing happens, you know, monthly or, you know, whatever that might be. So again, with the KPIs, it sort of will help us to define what type of tickets and the volumes, for example, of the customer inquiries that we want to automate, and how quickly do we actually want to be able to close that in automated fashion. And then the last point is really just to sort of adopt an iterative approach. And this is sort of akin to sort of using Agile where, you know, Deputy and really sort of iterate through continuous improvement. Go to the business, ask for feedback, look at your business needs, look at what changes in the business needs there are, how you can adapt your AI to meet those, and really just iterate through that. You're never going to start with the end solution. It's always going to be a journey. So I think with the iterations, keep people in the loop, Follow these points in the strategy, and I think that's when the value will surface by adapting or adopting conversational AI and adapting it specifically for your business. And of course, you could talk about pitfalls, Sascha. You can have your structure, but there's always going to be pitfalls that you need to look out for when you adopt this technology.

Yeah, absolutely. Yeah, just one point I wanted to mention, um, for the sort of— you talked about the iterative approach. Yeah, business improvement is, is, is never, never finished. So I think whatever it is, sometimes some people have that in their mind, I think, okay, we roll out this project and it's finished. But nowadays, um, when you, when you look at the business, yeah, you always have to look what's going on in the market. You have to adapt. Recent years have shown you always have to change, always. And that always applies here as well. So yeah, I think to avoid some pitfalls, yeah, it's super important to set some realistic expectations. We always hear like AI can do this and that and it saves the world and it can do everything, but it is really about realistic expectations. And starting small, like you mentioned earlier, is very important. And then understand really what you want to solve, and then you pick certain capabilities of AI. But AI as, as the one thing is not, it's not the silver bullet. And yeah, everyone needs to know really what AI can do. So this is back to this training you mentioned earlier, one of the strategies. So yeah, nearly is nearly the opposite of, of the successful strategies here. Um, yeah, and, and one, one big thing is, uh, data privacy. Everyone talks about that. Where, where are we using that AI? What happens with our data with AI? Is it in the public space of, let's say, OpenAI, um, copy and pasting information into ChatGPT or other things? Uh, or where is it? Where is that conversation there actually connected to? Um, what happens with that data? That's, that's obviously very, very important and that has to be carefully planned and you shouldn't allow that people to go crazy Wild Wild West style with it because that will come and bite you at the end and you have to have some really clear guidance internally around that and then of course there are some even some regulations around it. The EU AI Act, for example, you have to be very clear what are you doing with AI and data. So that should be a priority going forward. Yeah. And ensuring sort of cultural compatibility. You obviously have to see where you implement that, and this is why you're going small. If you do a big bang approach, then, and you say, now we have this great conversational business application available for everyone to use, you have to use it. That's part of the change management, I guess. And you have to make sure where you roll it out, people are actually excited about it, because only that, we learned that with lots and lots of these projects and programs. You need to get that excitement in, and only then the next department, the next application will be rolled out successfully. Then you can go bigger, and then suddenly the whole division is excited, and then you have this overall excitement. This is how change generally works. You just have to excite people about the benefits and about the change. Yeah, and, uh, uh, yeah, again, do it in small bits. I think we can say it, I think, not more often again. So do it really in phases, otherwise people get really sort of tired of it. So especially when projects really go along and on, and then in September we say, yeah, we're launching this new AI project, conversational AI, and then next September we're still working on it and only just a few people have seen it. So that, that happens in, in some organizations unfortunately. Um, yeah, avoid that, uh, by by having lots of phases and get people excited. Yeah, so yeah, embracing those kind of new technologies, conversational AI, conversational business applications, as we mentioned earlier, it really requires sort of careful planning. Sure, everyone should experiment, and that's nearly like Wild Wild West when it's nearly like in this experimentation mode. But after that, it needs sort of really careful planning, Transparency, especially when it comes to data and privacy and knowledge. Yeah, and focus on the people, as always. And, uh, yeah, if, if you foster that excitement, um, yeah, I think, um, yeah, you can go all the way with that. And it has to be a collaborative ecosystem, um, sort of there where technology and, and, uh, and people people work together. I think only then can it be successful. And yeah, I think that new style of applications we will see more and more often is very interesting. So I guess we still have to work a lot to convince people in organizations to go from applications, standard applications we have built over the years, so it's like a workflow application, and we have the apps and those kind of things. I think Some people go from spreadsheet now to apps, and if you now come with, hey, let's do build an application which is just 100% conversation, I think that's the hard bit as well to, to, to understand, to not understand, but to get that across where the benefits are as well, um, in, in the different departments. Um, but, uh, yeah, so I think that's something what we will see more and more often. Um And I'm excited to see it more often.

Yeah, no, I agree. I absolutely agree.

So yeah, with that, we are sort of at the end of the episode. Yeah, very exciting topic. If you'd like to know more about that, chatbots, conversational AI, virtual assistants, yeah, leave some feedback, reach out to us via our social channels, on our website theautomationguys.net, join our new community, the Automation Guys community. Lots of of good conversations going on there. Lots of sort of training sessions going on there on all these things. Yeah, join us there. And yeah, as always, leave us a 5-star rating, of course, wherever you listen to this podcast. And yeah, so we will be back with another episode very soon. And until then, let's automate it.