WP5 main objective is to ensure successful E2F2C framework delivery for distributed extreme-scale audio analytics. The framework allows for powerful, scalable and real-time processing of multimodal audio-visual data on top of distributed deployment of MARVEL ML models. In details, WP5 aims to ensure:

• provision and configuration of a powerful HPC infrastructure.
• orchestration of infrastructure resource management and optimized automatic usage of external computational and storage resources.
• seamless integration and quality assurance of software releases.
• quantifiable progress against societal, academic and industry validated benchmarks.
• real-life experimentation and validation at large scale.
• continuous alignment with the responsible AI planning and guidelines set out in the first phase of the project.