Microsoft Discovery GA: Agentic AI for R&D Unleashed

Microsoft is making its agentic AI platform for research and development generally available. This extends their AI strategy beyond workplace tools into scientific and engineering workflows. The platform is called Microsoft Discovery and was announced at Microsoft Build. It helps organizations build and govern AI-driven research across various fields, including materials science and life sciences. What's interesting is Microsoft also introduced a preview of the Microsoft Discovery app. This desktop experience is for researchers and students who might not need a full enterprise setup. The core idea is that AI agents should not just answer questions, but also plan, reason, and work through multi-step processes. This approach is applied to research and development, where teams often go through cycles of hypothesis and experimentation. The Microsoft Discovery Engine supports this scientific work loop. For organizations, governance is key. Microsoft says Discovery connects to institutional knowledge and simulation tools while keeping outputs reviewable and workflows reproducible. They emphasize keeping "human judgment" at the center of research decisions. The new Discovery app preview lowers the entry barrier. It can be downloaded from GitHub and used with a GitHub Copilot account. This allows smaller teams to explore literature review and hypothesis generation before moving to the broader platform. For example, Yale Engineering used the Discovery Engine for small molecule design in battery research. This matters because it shows how AI is being integrated into complex scientific discovery.