In this episode of the Process and Automation Podcast, hosts Speaker A and Arno explore how ChatGPT—a large language model based on the Generative Pre‑trained Transformer (GPT) architecture and fine‑tuned with Reinforcement Learning from Human Feedback (RLHF)—can be combined with Robotic Process Automation (RPA) to elevate enterprise automation. They begin with a concise explanation of what ChatGPT is, its training methodology, and its rapid rise as the fastest‑growing consumer app. The conversation then shifts to RPA’s role in automating repetitive, rule‑based tasks such as invoice and payroll processing. By integrating ChatGPT’s natural‑language capabilities with RPA’s structured workflow execution, businesses can create human‑like chatbots for customer service, transform unstructured data into structured formats for reporting, and even support advanced use cases like financial portfolio optimization. The hosts cite a PwC study showing 86 % of CEOs view AI as essential to daily operations, and they discuss how the pandemic accelerated digital transformation, creating pressure on IT teams to handle massive data flows. Throughout, they emphasize the complementary strengths of ChatGPT (creative text generation, language understanding) and RPA (speed, accuracy, rule execution), the endless possibilities this synergy unlocks, and the importance of community support for organizations just starting their AI‑automation journey. The episode concludes with a call to action for listeners to experiment with these technologies and stay tuned for future discussions.
Chapters
Welcome back everyone to another episode of the Process and Automation Podcast with the Automation Guys. Yeah, thanks a lot to all our community members for sending in plenty of questions regarding what is now a very hot topic, ChatGPT. And yeah, we received a couple of questions regarding ChatGPT and the use cases. Let's have a chat about exactly that today, ChatGPT and RPA. And yeah, how, how these two things combined are transforming the future of automation. So let's, let's go a bit deeper into that. Yeah, and as you know, in a couple of days generative AI has come to the forefront of the technology revolution. So it just appeared with ChatGPT. Yeah, and suddenly everyone was on it. Fastest growing app everywhere has enabled pretty much anyone to use AI to create books, articles, write code from scratch as well, come up with not with ChatGPT, but with other generative AI stuff, create fabulous pictures. And yeah, that's really, really amazing to witness because Arno and myself, we are of course trying to convince everyone to use these latest technologies, low-code, no-code workflow, AI for many years already. So now it has really got this huge attention. So yeah, and yeah, as I mentioned earlier, today we want to explore what happens when ChatGPT and RPA come together to, to really advance that revolution. So by integrating natural language processing, or in short NLP, into intelligent automation has become essential for lots of businesses out there, especially businesses who are keen to improve their efficiency. So with OpenAI's ChatGPT, we can now integrate ChatGPT with RPA. So by using that, we, we can really enhance the capabilities or enhance our automation capabilities. And Arno, do you like to go a bit deeper maybe into—
Yeah, sure.
For some of our listeners who maybe not just heard ChatGPT everywhere, but maybe haven't drilled in deeper to what it really is.
Yeah, yeah. Well, it's the obvious question about what is ChatGPT? And obviously not a day goes past where people actually talk about it. But ChatGPT is actually an acronym for Generative Pre-trained Transformer. That's what the GPT stands for. And it is an AI-based large language model, or LLM, and roughly launched about November last year. It's developed by OpenAI, and the dominant NLP technology used in ChatGPT makes it capable of generating human-like responses to conversational human queries. So deep natural networks are pre-trained on text data, and then it gets fine-tuned for tasks like text generation and answering questions. So, you know, with Insight, the technology, it uses a thing called reinforcement learning from human feedback. There's another acronym, RLHF. And, you know, this sort of reinforcing from the feedback is used for— it's effectively, it's training. So humans simulate conversations to train the algorithm. And then adapting the answers based on human dialogue to improve its accuracy over time. So it's a bit like a child learning a new skill. You know, it's repetitive. So today's topic, like you said, you know, we want to look at how ChatGPT and RPA really work together to augment each other. And of course, RPA, as we discussed in many previous episodes, is used to automate manual processes, you know, stuff that is repetitive in nature and things that humans do that you can ask a software bot to do more quickly and more accurately. So, you know, while RPA enhances process speed and things like, for instance, accuracy, for structured data, it can only automate repetitive logic-based tasks. So examples of automated processes includes things like, I don't know, invoice and payroll processing. So on the contrary, if you look at ChatGPT, you know, it can generate things like articles and poems and research papers and You said earlier it writes software code. It does this in seconds and, you know, it's sort of unbelievable to see it. So ChatGPT's potential to kind of generate natural human text without effort has, you know, made it a really, really hyped about technology. And it's, you know, the fastest growing software out there in human history. So Both ChatGPT and RPA obviously have their, has their shares of strengths. So, you know, this, this is why we, it'd be interesting to see, you know, how, how we can combine ChatGPT and RPA to work together to sort of take automation to the next level.
Yeah, yeah, it will be very interesting to see more and more of that and more of our community members using it. So yeah, using ChatGPT and RPA in that combination, yeah, brings it really to this new level. And now with ChatGPT interaction with these processes in a human-like way makes ChatGPT an ideal addition to workflows. So for example, with ChatGPT-powered chatbots, we can now integrate that technology in customer service processes to provide real-time responses. So obviously that's, that's what we heard before as well, but some of that stuff was really, really difficult to do. So now this human-like power of ChatGPT, that makes a real difference here, improving the customer satisfaction. And by using this or doing more automation, it will massively reduce the need for human labor.
Yeah.
Furthermore, doing this integration, ChatGPT and RPA, Yeah, we, we, we can dive deeper into the sort of data analysis and reporting tasks. ChatGPT, for example, can evaluate those kind of datasets and provide the outcomes. So it's very, very, very intelligent. Sometimes I'm amazed when you give it certain stuff, it comes, comes out with some really good interpretation of the data. And Yeah, it can therefore optimize customer portfolios, for example, when it comes to very specific financial institutions use cases. Yeah, so that's very powerful. And by combining ChatGPT with RPA, so as a business you can streamline the processes further. So Of course, that's always a driver for business looking at automation technologies, increasing the accuracy, reducing the cost, and of course, ultimately customer satisfaction and company performance. But yeah, combining NLP with RPA is sort of the next thing, and especially when it connects to your company data. So there will be endless possibilities for getting more efficient and more productive going forward.
Yeah, and I think that if you look at ChatGPT, it does— well, it's starting to play a sort of more integral role within cyber robotic process automation. Of course, you know, we, we look at the role of AI-based technologies out there in automating operations with inside businesses, and, you know, that's becoming increasingly valuable And AI has become really the driving force for automating business processes for businesses across various industries. So we look at automating within production floors using RPA to managing invoicing, inventory, providing customer support, and even things like you know, handling kind of queries within your HR department. And, you know, if you look at AI applications that, that's out there, it's really being used across a wide spectrum of functions, you know, within businesses. There was a recent study by PwC, and 86% of CEOs out there actually regard AI is an integral part of sort of their daily office operations or the ambition to bring it in. And, you know, digital transformation as we know it and, you know, and sort of the adoption has obviously accelerated in the last couple of years worldwide. You know, the COVID-19 pandemic forced a lot of organizations to obviously shift from, well, shift to digital operations, you know, growing their kind of online presence, being able to offer remote working. So this sort of sudden change put a lot of IT teams out there under immense pressure to manage, you know, a vast amount of data processes and in a lot of applications across multiple departments. So we feel that ChatGPT is a, is a perfect sort of complement to RPA in that with the NLP capability, it will provide a degree of turning these unstructured data into structured data. And that's essentially what we can plug into our RPA processes. So it's a very exciting addition to the automation stack. You know, it's something we talk a lot about on the show. I'm pretty sure in the future we're still going to talk a lot about all of the various sort of new capabilities in the various ChatGPT models that's being released. I know ChatGPT-4 is out at the minute. So yeah, it's a very an interesting space to sort of look at. And I think it's a practical way for people to really understand what AI is.
Great. Yeah, thanks a lot. Yeah, without a doubt, lots of potential really with these AI tools. And yeah, for everyone who is out there in our community, feel free to reach out to us. Yeah, so plenty of development when it comes to generative AI out there. And now combining all these different things in the right process, end-to-end, and plug it into the right data, your data in your company, I think was that. Yeah, as we mentioned it earlier, the potential is nearly endless. And I think lots of companies we had conversations with they're not even using the stuff which was hip 10 years ago, isn't it? So, but still, it is always difficult to get started and get your hands on these technologies to make meaningful change. So that's why we have this community here. This is why we're talking about this. Yeah, we are passionate about this topic. We will be back here on 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. Happy birthday!
