Radical AI

Working on machine learning and experimental systems for materials discovery.

We focus on speeding up how new materials are found and tested by combining modeling, data, and real-world experiments.

Materials research is still slow and expensive. It is hard to run enough experiments, and even harder to learn from them quickly. We are trying to tighten that loop so new ideas can be generated, tested, and improved much faster.

On this page, we share models, datasets, and tools related to that work.

Our work includes:

• property prediction and modeling
• dataset development
• experimental workflows

We care a lot about connecting these pieces so they inform each other instead of living in separate pipelines.