Secondmind

Secondmind builds AI software for engineers who can’t afford to test everything. Modern vehicle programs involve thousands of design variables and calibration parameters, and the conventional way to find the right ones is brute force: run the simulation again, book the test bench again. Secondmind’s active-learning system replaces that with a model that learns which experiment is worth running next, so teams converge on reliable results in far fewer runs.

The software covers two core workflows. Design Space Exploration maps the feasible boundary of a design, which the company says cuts CAE simulations by up to 90%. Calibration produces production-ready maps for engines, e-motors, and powertrains in significantly fewer test points. Both work with the simulation and test tools engineers already use and require no machine-learning expertise. The platform is ISO 27001-certified and cloud-native.