#71 Intelligent Automation, ChatGPT and some Use Cases

Welcome back to another episode of the Process and Automation Podcast, where we will explore all things process automation. So today we are discussing the importance of AI and natural language processing. And specifically in intelligent automation. So one of the topics we love to talk about a lot here. And again, specifically focusing on ChatGPT, a powerful language model developed by OpenAI. And I believe everyone listening in today has heard about ChatGPT. It's pretty much everywhere at the moment. So first, let's discuss what ChatGPT is and how it works. Arno, do you want to take that one?

Yeah, sure. So ChatGPT is like you say, it's, it's a new phenomenon. OpenAI created it. It's in beta and every time you look at it, there's just a million more subscribers. So it's the quickest service out there in terms of onboarding of new people using it. And essentially it's a machine learning model that uses a combination of deep learning techniques to generate human-like text. Now it's been trained on a massive dataset, you know, estimates out there say it's over 40 gigabytes of text. And this actually allows it to understand and generate, I guess, you know, a wide range of content in, in a particular language. So again, this, this makes it a very powerful and valuable tool you know, for businesses and organizations out there looking to automate tasks that involve NLP and natural language processing. You know, obviously things like customer service, services, content creation, things like language translations, contract reviews. And then, you know, one of the key advantages of ChatGPT is its ability to understand and respond to the context that you use when you talk to it. So this obviously allows it to generate text that's more natural and relevant, you know, to the type of task you wish to perform. So for example, if we look, I don't know, pick a customer service context. So ChatGPT can understand the customer's question and respond with a sort of a tailored accurate answer that's sort of in context with the conversation. And again, this greatly improved the customer experience and reduced the, the need for, for human intervention. Um, if I may, I'll just sort of just do one more advantage, and I'll, I'll give you some time to, to, to, um, provide some input as well. But, um, one of the advantages is obviously its ability to scale. Um, as we feed more data into the model, becomes better at understanding and generating language. So again, this, this means that the more business businesses out there that uses ChatGPT, the better it will become over time at automating tasks and improving things like customer experience.

Yeah. And Oh yeah, that's, that's really great, Arno. Yeah, and then in addition to that, ChatGPT can be obviously integrated with other technologies to create more powerful and efficient automations. So yeah, we're dealing with lots of systems within our customer projects, and this part of, or this AI tool can be plugged in like any other AI tool into the process. So Yeah, again, I have a couple of examples here. Yeah, it can be combined with even more natural language tools. So there's something out there called natural language understanding and natural language generation. And combining all these things, you could create a complete conversational AI system, really. And yeah, this can be used— I know you mentioned that already— for customer service, getting that customer experience experience improved, which is obviously substantial for businesses. It can be used for sales and other activities pretty much everywhere where you have human-like interactions. And yeah, now let's specifically look how a business can use ChatGPT to improve intelligent automation initiatives. So And I have to say the possibilities are really endless.

And yeah, it is, is really loads of examples.

Yeah, we could talk probably years about it. Similar like we, we talked about chatbot use cases in companies. And now, I mean, if you think about combining that obviously with ChatGPT, you obviously you are into the same use cases we talked about a lot here. So hundreds of those. Yeah. But yeah, let's Maybe one by one, Anna, we can go through some of the areas we researched. Yeah, automating customer service. Yeah, you can respond to customer questions and concerns so much better than we used to. And yeah, and then basically you will have less human customer service agents and needs in your customer service center that allows for 24/7, 365 days a year sort of support with really interactions. And that's what it is. We don't just want to have a sort of a form kind of chatbot. It needs to have that, that real understanding what's going on.

Yeah. Well, I guess that in a particular use case, there's not just kind of traditional customer services where I've got a faulty product and, you know, I'm speaking to an agent that helps me either troubleshoot it or help me process a return. I mean, you can think about use cases like triage. So patient triage, for example. In the sort of the medical sense where, you know, rather than somebody phoning in, they might sort of provide some of their symptoms and, you know, ChatGPT can actually learn from that and kind of make recommendations or escalate this to, you know, a particular clinician. So I think in terms of customer service, you know, if you look from a B2C, B2B, you know, like you say, you can, you can talk for months about the application of this. Then obviously the next one I've mentioned earlier, but you know, this is what they call sort of the Google killer is content creation. And you could use ChatGPT to generate high-quality content for websites and social media and, you know, various other platforms. And, you know, this again can save your business a lot of time and money. And, you know, it also allows you to produce a lot more content at a faster pace. So in all, I think that, you know, that is probably gonna be a big application of the technology. If I may, I'll grab the next one as well because I spoke earlier about language translation. But this is a very, very powerful capability that you can use to translate text from one language to another, making it, you know, very easy for businesses to communicate with customers and partners in different countries. You know, I guess gone are the days where it's sort of specialized language translators because ChatGPT has also got an application program interface. So it's essentially something you can integrate with your technology stack. So if you've got, you know, let's say your email systems and things like that, it can do the translation automatically behind the scenes. It's gonna be really, really powerful and kind of build those seamless communication, you know, abilities where the translation is sort of just done behind the scenes, you know, and that obviously saves a lot of time because nobody wants to copy and paste the email into Google Translate and then try and read it.

Exactly. Yeah. So the next one is language summary, summarization. So if you have lots of content as part of your processes to go through, ChatGPT can be used to summarize the large amount of text. And depending on what's in there, obviously you can then do certain actions in your processes, or when you are sort of a publishing house or so, maybe you can make use of those. There's lots of many reasons why you want to do a summary of a large text and Yeah, and then we come into the application of workflow automation. As we mentioned earlier, there is an API. ChatGPT can be integrated with workflow automation tools like you have already in-house to create natural language-based interfaces, for example, triggering and managing those workflows. Triage was one thing you mentioned earlier, Arno. So whatever is there, whatever the output is of that integration with ChatGPT, it can determine where where things go. And yeah, this can make it easier for all the users throughout the business to initiate and track tasks. Yeah, I think there's a huge, huge benefit there combining this with your workflow tool. Yeah, and then you see, I think everywhere as well, and we fiddling around with that was Power Apps or some whatever tool is out there, you can plug it in easily in no time.

So the next one I want to touch on is case management. And I think this is a very, very important use case as well, where you can use ChatGPT to integrate with case management tools, you know, to automate the triage and, you know, routing of customer inquiries and support requests to the right people. And again, the benefits there are obvious. You know, the first one that comes to mind is the the speed and accuracy of case resolution improves drastically. And that obviously is as a result of the reduction, or for the need for human intervention, is essentially.

Yeah, and I think it will continue in all the other areas. So we're talking about iPaaS solutions a lot, and of course it's very similar. So you can obviously use it there to improve your data extraction and data transformations and and putting data and loading data from different sources. And was that enrichment of the output of ChatGPT? So yeah, that continues as well with BPM tools. It will help to automate all the repetitive tasks around invoice processing or purchase order management, contract management, all those kind of topics. Can be, can be sort of improved on, on that side. Yes, that too.

Yeah. And then if I may, I'll sort of— there's another application of this which is really, really quite interesting, and it's to do with Integration Platform as a Service, or iPaaS. So just for our listeners out there, I think we did have a few podcasts on iPaaS. In the past, but essentially iPaaS is a suite of cloud services enabling development and execution and governance of integrated flows connecting, you know, any combination of on-premise and cloud-based processes, services, applications, data, you know, across your organization. So again, You can use ChatGPT, you know, to integrate with your iPaaS solutions to create automated data integration. Yeah, transformation processes. We all know data integration, data transformation, ETL, those things are really, really tough and can be very expensive. So again, you can bring this service into play. And again, you know, this can be used to automate the extraction, transformation, and the loading, you know, ETL of data from various sources, databases, APIs, flat files, XML files, emails, you know, the list goes on and on and on. So again, very, very exciting use case for—

because it can generate, isn't it? So you give it some output from whatever, in whatever format, and it transforms it just magically into the right CSV structure. And yeah, And off you go. Yeah, I think, yeah, a very, very important one we touched on earlier when it came to content creation as part of processes and lots of activities in businesses. We're still sending emails. I know it's not, it's not really the most sophisticated way, but it's one of the easiest ways of notifying people or communicating. But Yeah, so that's what you can use ChatGPT for as well, generating these kind of email content. You can create a bit more personalized email campaigns because of the power of ChatGPT. And of course, if you put out all on automation and that personalization, it will support lots of sales and customer support activities, incoming emails, and automatic reply, all sorts of stuff like that. And yeah, increase the efficiency of all the departments. Yeah, so that's email automation. And I guess we will see lots of those emails coming in from salespeople and other marketing people where this is fully automated. And yeah, well, it's exciting and at the same time. Yeah, so we probably have to train our, our email filters a little bit.

And then the next one is clearly chatbot integration. So you could take your existing chatbot platform and you can integrate ChatGPT with it. So you can create more sophisticated human-like conversational experiences, I suppose. I suppose this kind of comes hand in hand with the customer support, but customer service. So, you know, again, it's sort of that use case, but very important for sales as well. So it understands context, so you can train it to respond to sales queries and really any other tasks that involve human-like interactions.

And I guess, so when we talked about chatbot implementation and how to get started, so we're talking about obviously in use cases like fully end-to-end, sign up for an insurance, that kind of stuff, and select the right products. So lots of interaction there, lots of AI capabilities already needed in those kind of platforms. But now, or what we said back then, those kind of initiatives get better and better when your own AI starts learning your sort of your own business taxonomy and that kind of stuff. But now with the power of ChatGPT, and combining that in those platforms, lots of that can be— so it's far more, or you're much quicker, much quicker getting up and running because a lot of that understanding what you usually need internally first and to try and deploy all that stuff, that is already there and not 100%, but it's a huge improvement in that kind of area.

Yeah, absolutely. So the verdict overall, ChatGPT is really powerful. It's a tool out there. It, like I said, it's still in beta, but it can greatly improve obviously the efficiency and effectiveness of your intelligent automation initiatives. It is just another kind of tool in the armory that you could used to more, I guess, move more towards autonomous business processes. And as businesses out there continue to adopt AI and automation technologies, we can expect to see even more innovative use cases of this particular technology in the future. So very, very exciting. You know, we can't wait to see how this evolves into 2023. We will definitely keep an eye on this. And for our listeners out there, if this is all new to you, feel free to reach out to us. We've done research on the subject matter. We understand where it fits in. And, you know, it is one of those things where it will become mainstream very, very quickly.

Yeah, absolutely. And it's— there's no, no, no hour when you, when you scroll through social media where not someone is sharing specific insights how ChatGPT can be used in their industry, in their department, in their business. So that's very interesting to, to just follow. And every day you see like very creative usages of ChatGPT. I think that's That's where the power really is to figure out how you can use it for the business processes. And yeah, thanks for listening to this episode of the Process and Automation Podcast. We hope you found this very informative and valuable. If you like to give us feedback, head over to theautomationguys.net, leave us a message there, or connect with Arno and myself on LinkedIn. Thanks again for listening to this episode. We will be back for another episode very soon, and until then, let's automate it. Unfortunately, that's it again with this episode of the Process in Automation Podcast. If you like this episode, please give us a 5-star rating and don't forget to subscribe to this podcast so you don't miss any upcoming episode. We hope you will tune in next time, and until then, let's automate it!