Mantis Biotech is making ‘digital twins’ of humans to help solve medicine’s data availability problem

Mantis Biotech, based in New York, says it is building a solution to address gaps in medical data availability.

Its platform combines varied sources—textbooks, motion‑capture cameras, biometric sensors, training logs and medical imaging—to produce synthetic datasets.

An LLM‑based system routes, validates and synthesizes those streams, then a physics engine grounds the output and creates high‑fidelity renders used to train predictive models.

Those synthetic datasets are used to build physics‑based digital twins: predictive models of anatomy, physiology and behavior for aggregation and analysis.

Potential uses include testing procedures, training surgical robots, simulating medical issues or behaviors, and predicting injuries (for example, an NFL player's risk of an Achilles injury based on performance, load, diet and activity history).