Using Microsoft Discovery and AI Agents to Accelerate Low-Carbon Cement Additive Research

Construction materials account for about 40% of the global carbon footprint, with cement and concrete production as a major contributor. Saint-Gobain, described as the largest construction material producer worldwide, is seeking new additives and chemistry to increase cement strength and reduce carbon footprint. The discussion focuses on using Microsoft Discovery and Microsoft Foundry agents to build a knowledge base, connect simulation models, and generate new research hypotheses.

■ AI-supported research workflow
- Building a knowledge base from almost 100,000 articles and patents
- Connecting Microsoft Foundry agents with in-house simulation models
- Using prompts to orchestrate workflows, ask questions, and create hypotheses

■ Molecule evaluation and research speed
- Visualizing new molecules and their interactions within formulations
- Reducing knowledge-base creation from about a year to a few hours
- Increasing hypothesis output from one or two per month to five to six in six weeks

■ Industry context and sustainability goal
- Construction representing 40% of the global carbon footprint
- Cement and concrete production as a significant component
- Delivering sustainable solutions for construction materials

This content is suited to viewers interested in AI-assisted materials research, construction technology, and industrial sustainability. It provides an overview of how AI agents and knowledge bases are applied to cement additive R&D and what workflow changes are described. Viewers can gain context on the reported speed and hypothesis-generation improvements discussed for AI-supported research.