Risorius is a Korean AI for Science company focused on brain disease research founded in 2023. Brain disease is one of the hardest areas of medicine to study. Patients with the same diagnosis can have very different underlying diseases, and in a living patient, conditions cannot be changed and experiments cannot be repeated. Our goal is to give researchers what other fields of engineering have long had: the ability to simulate before testing in the real world. We call this the Virtual Brain.
At its core is a research agent. It takes a research question and works with a knowledge graph built from textbooks and the scientific literature, a world model trained on real patient data (currently in development), and the analytical tools available to AI today. We see this as a way to reduce the cost and time of clinical research. Predictions are not treated as proof. We validate them with partner wet labs and compare the real results with what the model predicted, and every result is fed back into the system.
Our work is grounded in real patient data. Over the past three years, we have carried out about 20 joint research projects, 15 of them IRB-approved, across epilepsy, psychiatry, sleep, neurodegeneration and multiple sclerosis. Our sleep research includes iRBD conversion to Parkinson's disease and Lewy body dementia, as well as narcolepsy subtypes. We also run four EMR-based hospital partnerships for prospective data collection, and we carry out de-identification ourselves within environments agreed with each hospital.
Our team brings together clinicians and neuroscientists, who define the research questions, with medical AI researchers and engineers, who build the models and the agent, and a partnerships team that connects this work with the institutions that need it.
We build AI research agents and multimodal patient models that answer hard clinical questions in brain disease, such as which patients will progress. We are looking to join consortia for HORIZON-HLTH-2027-02-TOOL-01 as the AI and modelling partner.
Destination 5: HORIZON-HLTH-2027-02-TOOL-01-two-stage: Development of predictive biomarkers of disease progression and treatment response by using AI methodologies for chronic non-communicable diseases