Large language models can revolutionize genomics research, streamline clinical documentation, and improve real-time diagnostics, but they struggle with rare diseases and unusual conditions where reliable data is scarce. Mantis Biotech is developing a platform that creates synthetic datasets to build “digital twins” of the human body, which can be used to study and test new medical procedures, train surgical robots, and simulate medical issues.
Mantis’ platform takes data from various sources such as textbooks, motion capture cameras, biometric sensors, and medical imaging. It then uses an LLM-based system to route, validate, and synthesize the data, and runs it through a physics engine to create high-fidelity renders of that dataset. This allows for predictive models to be trained, enabling the prediction of human performance in various scenarios.
The platform is particularly useful for edge cases such as rare diseases, where reliable data is hard to obtain due to ethical and regulatory constraints. Mantis’ founder and CEO Georgia Witchel aims to use these digital twins to test hypotheses about human behavior without exploiting patient data.
The startup has seen success in professional sports, including working with an NBA team to model high-performing athletes. Mantis recently raised $7.4 million in seed funding and plans to continue building out its technology and release the platform to the general public, targeting preventative healthcare and pharmaceutical labs.
Source: https://techcrunch.com/2026/03/30/mantis-biotech-is-making-digital-twins-of-humans-to-help-solve-medicines-data-availability-problem