Provided and advised by: Henryk Mustroph

The EU Artificial Intelligence Act (EU AIA) [1] mandates that providers and deployers of high-risk AI systems establish a quality management system (QMS). Among other criteria, a QMS shall help verify and document the AI system design and quality. Current research rarely explores practical solutions for implementing the EU AIA. Instead, it tends to focus on theoretical concepts. As a result, more attention must be paid to tools that help humans actively check and document AI systems and orchestrate the implementation of all legal requirements. We implemented an initial draft of a web-based QMS that follows a microservice architecture and translates the requirements of the EU AIA into software requirements for standalone sub-service components within the QMS AIA [2]. The idea of this design is shown in the figure below. The goal is to provide the QMS as a Software-as-a-Service (SaaS) web application composed of multiple sub-services that operate as independent systems. These sub- services are organized into three core phases of the AI system lifecycle: pre- development, post-development and pre-release, and post-release. Each sub-service can be designed independently to implement specific EU AIA requirements for high-risk AI systems.