Artificial Intelligence GenAI in ARIS

In medicine, AI aids in recommending treatments based on patient data and predicting health outcomes, which helps in providing quicker and more accurate care. In marketing, it uses data-driven insights to identify target customers and create tailored experiences for them. In software product development, it automates and speeds up lot's of routine processes, from writing code to automating testing. In manufacturing and services, AI enables efficient research and the implementation of innovations.

Artificial intelligence has been introduced in ARIS Connect 10.0.26 (SR26). It is implemented using GenAI technology, which stands for Generative Artificial Intelligence. GenAI a neural network that generates new content based on the information it was trained on. This content can include images, texts, videos, audios, and other similar formats.

The application using GenAI technology and named ARIS AI Companion is integrated into ARIS Connect 10.0.26. ARIS AI Companion is a companion program that features a special language model and allows users to receive responses from artificial intelligence in a chat format to their verbal requests. Currently, requests can be made in the English and German languages.

There are three ways to use GenAI in ARIS Connect:

Model generation based on input text

The text may be a copied regulation or the result of an interview with a business expert. Artificial intelligence will create a process model in a specific notation based on this text (Fig. 1). Currently, process models can be generated in BPMN notation.

Figure 1. An example of creating a model in BPMN notation based on a text description

If necessary, you can edit the model generated by AI Companion, fill in attributes, and work with it as you would with any other model in ARIS.

It should be noted that when generating a model, GenAI does not use occurrences of existing object definitions available in the database but creates new ones. Therefore, users should consider this feature and subsequently integrate the new object definitions into the existing database.

Searching for information in a database based on input text

Artificial intelligence in ARIS can help find the needed information in a database. For example, by entering the name of a position or role, you can receive information about where this position or role is located in the organizational structure, who it reports to, and who it manages (Fig. 2).

Figure 2. An example of a direct response to the simple request

ARIS AI Companion is able to provide not only direct answers to simple requests, but also generate more complex search results for less specific questions. For example, if you search for "how to reduce the risk of fire", the system will find the relevant models in the database, which will show the necessary initiatives to reduce risk, as well as indicators measuring risk and responsible employees for them (Fig. 3).

Figure 3. An example of a comprehensive response to the general request

With the help of smart search, users can test various hypotheses about the behavior of the process, identify its bottlenecks, or search for opportunities to improve the process.

Assistance to business users in interpreting data obtained in ARIS Process Intelligence

GenAI technology supports users in analyzing data obtained from information systems where business processes occur. The Process Intelligence component provides various process-related data, presenting it in the form of statistical samples, graphs, and diagrams. Previously, users needed statistical analysis tools to interpret this information. Now, the AI Companion component analyzes this data and, using a query entered in a natural language way, provides users with the resulting information on the correlation of key performance indicators (KPIs), offers suggestions for optimizing processes, and helps generate custom calculated fields to measure KPIs.

Thus, the following advantages of using GenAI technology in ARIS Connect can be highlighted:

1. Saving time when creating a model. The model is generated within seconds.
2. The ability to make decisions about optimizing bottlenecks based on the processed information.
3. The ability to quickly and intelligently search for information in the repository.
4. Saving time on analyzing the correlations of various business process indicators.
5. Quick and informed definition of new estimated KPIs or correction of existing ones.
6. Better usage of Process Intelligence results.

However, there are some disadvantages to using GenAI technology due to the novelty of the product:

1. Currently, ARIS AI Companion is only included in ARIS Cloud Enterprise license.
2. AI Companion supports only English and German languages.
3. When generating the model, only definitions of objects are created and users have to implement new definitions in the existing repository.
4. Currently, it is not clear what other types of process models besides BPMN can be generated automatically based on the text information.

By Q4 2024, only the pilot version of ARIS AI Companion with reduced functionalities was available. The new version, with all functionalities and enhancements, will be available in 2025.

More detailed information about GenAI and ARIS AI Companion is available on the official website of Software GmbH (form. Software AG) or in the ARIS community.

If you have any questions, please, contact us by the contact form or at info@dainovaro.com and we provide you with more information.