The platform features a highly extensible architecture, allowing integration of partner, open-source, and proprietary models, tools, and datasets
Its core is a graph-based knowledge engine that enables deep, contextual reasoning over complex, multidisciplinary scientific data
Specialized AI agents, defined in natural language, collaborate with researchers to automate and optimize tasks like simulation, literature review, and hypothesis generation
Microsoft Copilot acts as the orchestrator, coordinating agents and workflows, and integrating AI and HPC simulations for end-to-end R&D processes
Discovery supports iterative, evidence-driven research, adapting strategies based on findings and enabling transparent, traceable reasoning
Built on Azure, the platform ensures enterprise-grade governance, compliance, and security, while supporting open innovation and future technologies like quantum computing
Real-world impact includes rapid discovery of a non-PFAS coolant prototype for data centers and a solid-state electrolyte candidate using 70% less lithium