Eric Bax

Secondmind develops Bayesian machine learning optimisation technology for engineering applications, with a primary focus on automotive powertrain and vehicle systems calibration. The Cambridge company builds probabilistic machine learning tools that help automotive engineers optimise complex, multi-variable systems such as engine management, transmission control, and electrification systems with fewer physical experiments and less calibration time than traditional approaches.\n\nThe company raised approximately $25 million in a Series A round and acquired Prowler.io, a multi-agent reinforcement learning startup from Cambridge, in 2021 to expand its AI research capabilities. Secondmind customers include major automotive OEMs using the platform to reduce the number of dyno tests required for powertrain calibration, translating directly into reduced development costs and faster time-to-production for new vehicle programmes.\n\nSeconmind operates in the engineering AI optimisation market, competing with specialist automotive AI vendors and general-purpose Bayesian optimisation tools. The automotive calibration market is driven by increasing powertrain complexity as the industry transitions to hybrid and electric architectures, which require calibrating interactions between internal combustion engines, electric motors, and battery management systems that generate dramatically more control parameters than traditional powertrains. Cambridge University research partnerships provide the company a continuing source of methodological development in Bayesian and probabilistic machine learning.