Biography: Tian is a principal research manager and project lead at Microsoft Research AI for Science. He lead a highly interdisciplinary team of researchers, engineers, and program managers to develop foundational AI capabilities to accelerate the design of novel materials, aiming to impact broad areas including energy storage, carbon capture, and catalysis. He also lead the development of MatterGen, an AI generator that discovers novel materials. His team also develops MatterSim, an AI emulator that accelerates the simulation of material properties.

Before Microsoft, Tian was a postdoc in the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT from 2020 to 2022, co-advised by Tommi Jaakkola and Regina Barzilay. He got his PhD in Materials Science and Engineering at MIT in 2020, advised by Jeffrey C. Grossman. Tian also did research internships at DeepMind and Google X.

Tian’s most noticeable work before Microsoft includes the development of CDVAE in 2021, a generative model for materials that significantly surpassing other models at the time, as well as CGCNN in 2018, the first graph neural network specifically designed for materials.

MultimodalAI'25 Keynote Title: Accelerating materials design with AI emulators and generators.

MultimodalAI'25 Keynote Abstract: The design of novel materials has been a cornerstone of technological progress, driving transformative innovations such as the adoption of electric vehicles, the development of highly efficient solar cells, and the widespread use of superconductors in magnetic resonance imaging (MRI) systems. At Microsoft Research, we develop two foundational artificial intelligence (AI) models to accelerate the materials discovery process. The first model, MatterGen, is an AI generator that proposes novel materials candidates given prompts of required properties. The second model, MatterSim, is an AI emulator that then simulates the properties of the generated candidates for the target application. The two models work together as a flywheel to drive the discovery of novel materials for broad applications. This presentation will provide a comprehensive overview of the architecture MatterSim and MatterGen, as well as how they can be used to deliver real-world impact in materials design.