Foundations and Collaborations

A deepening partnership with NVIDIA

TimeTrace Labs and NVIDIA have entered a long-term collaboration, applying world model architectures to physiological time-series data. Sonata is the first output: a world model of human motion, co-developed with NVIDIA researchers, published and available now . More will follow soon.

The work reflects a shared view that measurement, not model scale, is the binding constraint on progress in healthcare AI.

UK Government support

TimeTrace has been awarded two forms of non-dilutive UK Government support for this work: an Innovate UK Frontier AI grant, and an AIRR allocation providing compute on Isambard AI, the UK's national AI supercomputing infrastructure. Together they give us the funding and the compute headroom to push our world model work, and the TimeTrace platform, further and faster.

Academic and data partnerships

Talent

We run a structured internship programme with the University of Cambridge's Centre for Doctoral Training in Data Intensive Science, bringing mathematicians, physicists and machine learning engineers onto the measurement platform.

A separate pipeline runs into the University of Oxford, where the company began and where our decade of proprietary longitudinal data originates.

ETH Zurich

We are collaborating with ETH on efficient world models for complex physiological and multimodal time-series data, through co-funded postgraduate projects beginning in September 2026. The work targets low parameter footprint and edge-compliant deployment, so that models trained at national scale can run on the device holding the data.

Fondation FondaMental

FondaMental is a French scientific foundation created at the initiative of the Ministry of Higher Education and Research. It curates longitudinal mental health cohort data gathered across France at exceptional scale and duration.

We are applying the platform directly to that dataset, validating it against real clinical trajectories at national scale.