Rotary Club de Paris names Hydrobio Startup of the Year

The Rotary Club de Paris has named Hydrobio its Startup of the Year 2026, recognising the formulation’s technical novelty and impact potential.

A jury from outside the sector

Most of the prizes a company like ours competes for are judged by people who already work in agriculture or in venture capital. They know the vocabulary — cation-exchange capacity, mycorrhizal colonisation, biostimulant regulation — and they rank you against other companies making adjacent promises. That is useful, and it is also a closed loop.

The Rotary Club de Paris is a civic service organisation, not an agtech network. A jury drawn from it evaluates a soil amendment the way anyone outside the sector would: does the problem actually matter, is the mechanism real, and would it change anything for the people living with it. Those questions are harder than they look, because none of them can be answered with sector jargon.

The problem it was judging

A degraded soil fails physically before it fails chemically. Structure collapses, so the soil no longer holds water; water that is not held drains, and takes dissolved nutrients with it; nutrients that leach get replaced with more fertiliser, which leaches in turn. Fertiliser treats the symptom. It does nothing about retention, which is the actual fault.

Hydrobio works on retention first. A clay-based mineral matrix, micronised to a controlled granulometry, exposes far more surface to the soil than the same material as coarse grains — roughly two hundred times more. That surface is what holds water and what carries the cation exchange that keeps nutrients in the root zone instead of below it.

The second half is biological. A living mycorrhizal consortium travels in the matrix, and its hyphae extend the effective reach of the root system well past the root hairs themselves. The minerals give the fungi a physical scaffold to colonise, which is why colonisation is faster and why the effect is designed to last three to five years from a single application rather than a season.

What "technical novelty" refers to

Mineral amendments exist. Mycorrhizal inoculants exist. The claim in Hydrobio is the combination — a mineral matrix that functions as a carrier and habitat for a living consortium rather than a bag of two things mixed at the last minute. A USPTO patent covering the composition, the mechanism of action and the process has been filed, with a PCT extension underway. Filed, not granted.

The second novelty is that there is no single product. HydrobioFormul is a supervised machine-learning engine: a soil analysis, a target crop and a climate context go in, and the exact mineral-to-mycorrhiza ratio and per-hectare dose come out, with the reasoning documented. It does not generate text or reason in the abstract. It learns numerical patterns from soil analyses, pedoclimatic data and trial outcomes, and every new trial improves it. A grower does not buy a sack. They get a recipe for their parcel.

What "impact potential" has to survive

The honest position is that Hydrobio has strong numbers from controlled conditions and a limited number of seasons in the ground. Controlled trials tell you a mechanism works. They do not tell you it survives a farm — a wet spring, an uneven spreader, a soil nobody characterised properly, a grower with other things to do that week. Only field trials that we neither design nor host will settle that.

So the next twelve months are unglamorous. Pilot parcels across a wider range of soil types, with water use, yield and soil indicators measured against untreated controls on the same farm. The EU 2019/1009 homologation dossier, which is in preparation, has to be completed before broader French commercialisation from late 2026. Each returned field result also feeds HydrobioFormul, which is the compounding part: the engine gets more accurate exactly as the evidence base grows.

A jury outside the agtech bubble judged the problem worth solving. Whether we solved it is a question for the soil.