OpenBrain's platform is based on Baysig, a new statistical programming language. Baysig combines general-purpose computing with flexible statistical modeling for a wide range of data types. In Baysig, we program directly with probability distributions and dynamical systems and use Bayesian inference to link models with real-world data.

  • Bayesian statistics

    Provides the link between statistical models and real-world data and can quantify uncertainty and inform optimal decisions.

  • Functional programming

    Mediates the combination of statistical inference and general-purpose computing so we can go all the way from raw data to decisions.

  • Dynamical systems

    in the form of ordinary or stochastic differential equations are powerful models that describe how systems can change in time and forecast their future states based on historical data.

  • Typed data

    We have developed a uniquely powerful representation of complex and heterogeneous real-world data based on type theory.

Further reading: Quick tour of the Baysig language | Baysig reference manual