A Great Combination

One of the most common questions I have been asking to myself lately is what would be a good use case combining RPA and LLM as the term LLM (Large Language Models).

Lately, when we hear the term "LLM", we only think of ChatGPT. Even though the ChatGPT is based on LLM, it is not the only solution using an LLM.

The Communications Mining at UiPath platform, for instance, is using a state-of-the-art LLM to automate the reading and understanding of business conversations. It uses an unsupervised learning for clustering and then with the help of the supervised learning, it helps, for instance, classifying unstructured data, like emails, tickets, etc. And it is not using ChatGPT !

But in my demo, I decided to use ChatGPT. I combined a chatbot, an unattended robot and the ChatGPT LLM and created the following simple automation.

I should also clearly mention that it is not applicable to the real world as it is now, but I hope, it can be a good base to more complex, real-world relevant automations.

Here comes the scenario, after which I will go into some technical details:

 

• A doctor meets a patient and writes down the symptoms based on the conversation.
• He never comes to the point to give the diagnosis to the patient.
• The patient then is referred to a chatbot to question his symptoms to get some diagnosis.

Let's have a look at the video below:

 

As the first step, the patient provides his account id and this is verified by the unattended robot by querying the database, which is the excel sheet. The robot then retrieves the symptoms based on the account id, passes them to the ChatGPT via an API call and receives the response back with the possible diagnosis and sends them back to the patient via the chatbot. While doing that, it also sends an email with the same information to the patient's email address which was also retrieved while querying the account id.

To me, what is the most fascinating is the possibility to use a front-end AI enabled  chatbot, with an unattended bot and a back-end AI chatbot built on Generative Pre-trained Transformer language model. AI is a great technology, but it always needs a robot to make things done, like querying a database or sending an email with information and while building automations, this should always be taken into consideration. A brain always needs hands and feet to control !

Please share your ideas on how this simple case can be turned into a more productive one to create more value...

Think RPA and AI together

It has been some time since I published my last blog entry, which I am aware of. As of this year, my role at UiPath has changed to focus only in AI products of the platform and due to this reason, I wanted to revive the blog. I will be sharing some use cases, my views on AI in general and also talking about new AI functionalities/capabilities of the UiPath Business Platform in the coming weeks. Stay tuned. Hope you like my first article below. Please share your views. 

AI is not a hype anymore, it is everywhere. It is not anymore a technology of the future, but instead we are surrounded by it. Look what happened with Chatgpt, which was officially released on 30th of November 2022. It crossed the 100 million users milestone in January 2023 and in the first of month of its launch, it had more than 57 million monthly users. It was a real game changer and a great step forward to reach AGI, Artificial General Intelligence. And we can all be sure that more to come.

RPA and AI are also very close-to-each-other technologies (I call it RPAI) and when used together, they bring huge benefits to the business.

Looking at the UiPath Business Automation Platform, we can easily see that it embraces AI in different ways to meet the different business challenges. Studio, for instance, uses AI capabilities in an embedded way, like suggesting the next-best-activity as well as enabling the computer vision technology which is used while recording the actions before converting them into the activities. Simply put, you can create a whole RPA flow thanks to the computer vision without being in need of creating it manually, one by one.

The same applies for the Test Manager as well since it can also use the computer vision technology while creating the test cases. This of course applies even to Task Mining, especially the Unassisted one, which uses an ML model to identify the repetitive tasks to further explore and reveal the automation and process optimization opportunities.

As process automation expands to meet the needs of digital transformation and digitization initiatives, the ability to rapidly discover and analyze existing processes objectively and at scale becomes an imperative according to Forrester. UiPath Task Mining exists as part of the platform to embrace this challenge. It uses a data driven approach to gain deeper understanding of existing business processes happening on employees’ desktops to identify process improvement areas and automation candidates. It helps the business accelerate with the automation pipeline and discover by mining the unknown areas to identify repetitive tasks with AI-powered analysis by using a ML model deployed in AI Center and gives you a clear picture on the extracted data.

When talking of AI and RPA, we should also mention AI Center which is a very important component of the platform. It helps the businesses orchestrate all the moving pieces of AI, like deploy, consume, manage and improve machine learning models. It really bridges the gap between RPA and the data science teams and enables you to instantly apply the limitless cognitive power of AI to any software currently being automated by RPA. AI Center includes some ready made ML models (we call them OOB (Out of the box) models), which you can use and train according to your needs. You can also bring in your own ML models and deploy them. You can think of AI Center as a container of the ML models, some of which are also used by the different components of the platform like Task Mining and Document Understanding.

I would also like to elaborate more on Document Understanding (DU), after having mentioned about it. There is no company that is not dealing with the documents in different ways. But mainly, with the trapped information in the documents, there are some challenges in finding and extracting them. Processing documents is a challenge, especially if the work is done manually. It is prone to human error, it is pretty repetitive and sometimes (if not always) even boring. An automated DU powered by AI can wipe out those challenges and lead to cost and time efficiency by removing the risk of making mistakes, which might even have big consequences. In a DU framework, you can also involve in humans to verify the findings and train the ML models with their inputs to become smarter. AI plays an important role within those processes. Let’s consider invoices and assume you want to extract the invoice number, the total sum and the company name from them. As we all know, there is no two identical invoices in terms of templates since there are different suppliers that send those invoices and they have of course the freedom to create their own invoice templates. 

An ML model, deployed via AI Center, can help in identifying those fields regardless of where they are located in those documents. With the time, ML models become better and learn where to look at for those fields regardless of the template and even the language the invoices are created in. 

Classifying emails, is also a huge task, especially when tons of them come in, for instance to the customer helpdesks. Managing this manually costs too much, both in terms of money and time. Once

classified, moving them to the right folders is even a bigger challenge since there might be different rules applying to different folders. AI can also help with this challenge. A subset of AI; NLP (Natural Language Processing) is the technology enabling machines to understand human written text by using AI. With the acquisition of Re:infer, UiPath added a very important component to the platform to achieve this. The Re:infer platform is now integrated with the UiPath platform and has also been rebranded as Communications Mining, which you can read more about here.

All in all, AI is a great helper to automation when it comes to adding more intelligence, but we should also consider the execution, which is similarly important. You can extract information from the invoices, you can classify those emails, but when it comes to moving the extracted information from those invoices to a CRM system or moving those classified emails into the correct folders, you need RPA acting as hands and feet.

To conclude, I would like to refer to a Gartner analysis from 2019, where they mentioned that by 2022, 80% of organizations that deployed RPA would introduce artificial intelligence (AI), including machine learning and natural language processing algorithms for improving business processing activities. I am sure this transition will have a good momentum during this year and onwards with the introduction of the generative AI models as there will be more use cases to combine them both to achieve RPAI. This is where the real strength of the intelligent automation surfaces.

Think RPA and AI together ! Think RPAI !