Development of modern information solutions in the low-code platform oracle apex on the example of enhancing the funcionality of an existing application

This thesis explores the development of modern information solutions in the low-code development platform Oracle Apex, with a focus on enhancing an existing centralized identity and access management (IAM) solution.
In the theoretical part, we introduce the concept of low-code platforms, highlighting Oracle Apex, which is based on the Oracle relational database and uses the SQL and PL/SQL programming languages. We describe its core features, architecture, functioning and user interface. We also examined the role of artificial intelligence (AI) in the development of information systems in low-code environments.
We analyzed key aspects of IAM systems, highlighting key business processes supported by this system. In addition, we discussed the importance of audit trails for ensuring traceability and accountability. We analyzed the existing central IAM solution within the Oracle Apex environment, focusing on its structure and core business processes.

We integrated audit trails into a selected business process by extending the existing software library. Furthermore, we implemented two generative AI services, Cohere and OpenAI, into the low-code environment and added an AI assistant that generates queries and answers in natural language regarding audit trails based on a system prompt. The responses of the selected services were then compared.
During the upgrade of the information solution, we found that low-code development environments enable rapid development of simpler solutions or application prototypes without requiring programming knowledge. However, more complex upgrades require knowledge of programming languages.

Specifically, in our case, knowledge of SQL and PL/SQL languages.
We confirmed that the central IAM solution enables transparent and controlled access to functionalities in information systems. We also confirmed that historical tables alone do not provide sufficient insight into the business processes. After implementing generative AI solutions, we evaluated them using performance criteria (accuracy/quality, capacity/complexity and resource usage).

We found that the OpenAI service outperformed Cohere service in the context of this thesis.