We’re building an AI system for analysts and scientists, based on a fundamentally new approach to reasoning and knowledge representation. Our approach differs from LLMs in that we compose algorithms symbolically to represent complex knowledge, and perform probabilistic computations. This enables the AI-driven application of statistical models to different problems, while providing the user with a verifiable reasoning path, and an assessment of the uncertainty in each answer. We are developing applications for analysis and research in domains such as Finance, Strategy Consulting, Engineering, Material Sciences, and more.