Godela

Godela is a physics AI platform that trains models on existing simulation data to learn the governing behavior of physical systems. It addresses the time and compute cost of running repeated solver simulations by letting users ask what-if questions and evaluate configurations against physics-constrained predictions in seconds rather than days.

The workflow has three stages: upload simulation data to train a physics-constrained model, explore the system by changing parameters such as airflow, geometry, and materials, and optimize configurations against a chosen objective. The platform supports domains including data center thermals, electronics cooling, aerodynamics, electromagnetics, and structural analysis, and can build models from a question, geometry, or scenario. The homepage presents a data center cooling example with thermal distribution visualization, configuration ranking, and airflow redistribution.

Godela is aimed at engineering and R&D teams working on physical system design. Access is offered through a pilot request rather than a self-serve plan; no free tier, trial, or seat-based pricing is stated.