This deliverable presents advancements in evaluating recycled materials for high-performance construction, integrating these materials into sustainable building practices and supporting circular-economy goals. It introduces a large open-science dataset combining decades of laboratory tests with carbon-estimation methodologies, enabling prioritization of sustainable materials, particularly for concrete with reduced embodied CO₂. The core work experimentally and computationally validates new AI-driven design methods—adaptive design and zero-shot design—minimizing laboratory testing by using artificial intelligence to accelerate development of complex sustainable formulations. A major output is the SLAMD software, a digital laboratory twin reaching TRL-6, tested under realistic conditions to optimize material compositions to meet ecological and economic criteria. The report includes six case studies demonstrating application across Europe, such as recycled aggregates, sustainable asphalts, secondary cementitious materials, and mid-scale construction sites. Together, these results show the potential of AI-assisted design to transform sustainable construction practices.