The Foundation for Innovation and Technology (FIT) announced, in a press release dated 1 September 2026, the award of five new Tech Grants of CHF 100,000 each to projects in the canton of Vaud. Three come from CHUV (Chorus Tre, HeteroSight and SixtineAI) and two from EPFL (Cookd Labs and Endobotica). In total, this amounts to CHF 500,000 shared among five teams, all still at the project or prototype stage.

The CHUV projects are supported by a FIT-InnoTREK grant, the EPFL ones by a FIT-Innogrant. In each case, the foundation names a project lead and specifies what the money will be used for. It is these steps, more than the technological promises, that reveal how mature each project is.

Three CHUV projects focused on health data

Chorus Tre is developing a secure platform that allows hospitals and researchers to analyse sensitive medical data without it leaving their protected environment. According to FIT, the tool cuts the time needed to access this data for research “from several months to a few days”. The grant is intended to enable Vanessia Wu to commission an independent security audit, launch initial pilot projects with partner hospitals and set up a spin-off. The audit has therefore not yet taken place: at this stage, the announced time saving is a claim made by the project.

HeteroSight is designing an artificial intelligence tool that analyses a standard biopsy slide to determine whether a tumour contains several genetically distinct areas. The clinical stakes, as described by the foundation: some parts of a tumour may respond differently to a targeted therapy, and this information would help doctors choose the treatment. The funding will allow Sadegh Saghafinia to build a first patient cohort and validate the model’s ability to predict this heterogeneity from biopsy images. In other words, the predictive capability is the hypothesis the project still has to demonstrate.

SixtineAI is developing a platform for hospitals, clinics and independent doctors to deploy and govern clinical AI agents while keeping control of their sensitive data. Connected to existing medical systems and patient records, these agents are meant to assist doctors, nursing staff and other professionals with clinical and administrative tasks, such as drafting discharge letters. The grant is intended to enable Sami Perrin to launch the first clinical pilots at CHUV, validate the platform’s clinical benefit and safety, prepare its certification as a medical device and set up a spin-off. Certification is therefore a step still to come, not a given.

Two EPFL projects between robotics and AI

Cookd Labs combines a robotic laboratory and artificial intelligence to develop, test and improve food formulations more quickly, for example to replace certain ingredients or to create healthier, more sustainable and more affordable alternatives. According to the press release, what used to take several months of manual testing could be done in a few weeks. The grant will support Germain Meyer in validating the robotic prototype and developing a first AI product for food manufacturers.

Endobotica is developing TREX, a teleoperated robotic system designed for the treatment of ischaemic stroke. During this procedure, a doctor removes the clot blocking an artery in the brain; TREX is meant to help locate it better and remove it with greater precision and safety. The funding will allow Evgenia Roussinova to demonstrate the system’s reliability in the laboratory and improve it based on feedback from future users. The project is therefore at the stage of laboratory testing, not clinical use.

What is established, what remains to be proven

The established fact is the award of the five grants and their amount. The performance figures mentioned (months reduced to days or weeks, greater precision in the procedure, prediction of tumour heterogeneity) are goals or claims made by the teams and relayed by the foundation. The press release cites no scientific publication or quantified validation results in support. This is, in fact, precisely what the grants are meant to fund: audit, patient cohort, laboratory demonstration, clinical pilots.

From research to start-up

What the five projects have in common is their pivotal position, between the laboratory or hospital and a potential company. Two of them, Chorus Tre and SixtineAI, explicitly mention setting up a spin-off among the funded objectives. The others are first aiming for a technical or clinical validation step, a prerequisite for any commercialisation. Cookd Labs mentions a first product for the food industry, with no timeline specified.

For research in Vaud, these grants therefore target a specific phase: the one in which an idea born at CHUV or EPFL must be shown to work outside the research setting before a company can be considered. There is no guarantee that all five projects will clear this hurdle. The results of the announced pilots, audit and validations will show which of them are in a position to become companies.

Sources