<figure class="alignright size-full is-resized"></figure>

Curro Polo is currently pursuing an industrial PhD in the Basque Culinary Center program, with the academic guidance of Harvard University. His research project is taking place at Ama Brewery, where he is on the R&D team with Chef Ramón Perisé Moré who we interviewed recently. Curro is exploring the fascinating world of microbiology as it applies to beverages. We’d met at the 2024 KBI conference in Reno, Nevada, where Curro and I were both Kombucha Kup judges. At that time, he was finishing up his Master’s in Gastronomy and Culinary Arts at Harvard. I reported on his Master’s thesis, Kombucha: A Word on Metamorphosis, in March 2025.

Curro has launched a campaign to fund the first experiments of his PhD project: Open Flavor: Modeling Fermentation Through Open Science. The goal is to scale and model fermentation processes for the No/Lo (low and no alcohol) beverage category using open science and open source tools. The idea is to rethink fermentation beyond traditional beer and wine, exploring substrates like teas, and to make all methods, data, and tools openly available. He will utilize platforms such as the Pioreactor, an open source mini bioreactor. Expanding the cluster will enable him to conduct stronger, reproducible experiments that benefit both science and industry.

The Pioreactor

<figure class="alignleft size-full is-resized"></figure>

The Pioreactor is an affordable and user-friendly bioreactor that is flexible and easy to scale. It’s an open source tool for controlled micro-fermentations where variables such as pH, oxygen, temperature, and agitation can be precisely managed. Micro-fermentations of 20ml not only provide fast, cost-effective, and reproducible data generation but also create a foundation for predictive models that can be scaled to industrial production. This approach brings both scientific rigor and competitive advantages to the No/Lo industry.

A bioreactor is basically a machine that helps you to create an environment where you can control temperature, the movement, or whatever liquid is inside. You can also control what kind of other liquids you would put in the liquid for a pumping system. And then it also has a system, which is called optical density, that it measures how fast your cells are growing in that liquid, right? Because there’s many ways to approach fermentation, right? How I like to imagine fermentation is like a system, there’s a box and in that box you have a liquid, and that liquid has different compounds.

Initial work will focus on three non-Saccharomyces yeasts relevant for No/Lo beverages. We will systematically vary temperature, pH, oxygen, stirring speed, and tea substrate to assess strain performance. Growth curves and CO₂ production, measured directly in the Pioreactor, will provide high-resolution insights into metabolic activity and fermentation dynamics. These datasets will support the first generation of predictive models, later validated against 500 L and 2000 L fermentations to test reproducibility and scalability. Positive and negative controls will anchor results, ensuring the models capture strain-specific behavior and guide cost-effective, science-based innovation in the No/Lo sector. The fermentation is studied in 40 ml vials, where 10-30 ml of liquid ensures the liquid-air interface is optimized and tubing remains clear of the liquid.

Crowdfunding success!

Curro reached his goal of $2,564 on September 22, allowing him to expand the current cluster from nine Pioreactors to 15, and conduct experiments with five variables in triplicate.

<figure class="aligncenter size-full is-resized"></figure>

Here’s his description of the project:

The whole idea of the project is not to create a product but rather to test a statistical model. You can think of any mixed fermentation as a party where different guests are wearing unique perfumes. If I had a sensor that could detect each perfume, I could assume which guest is doing what and maybe even predict their behavior. I will give you an example with a very reduced consortia: Imagine you have a sweet tea with a yeast strain (that produces CO₂) and also a bacterial strain (let’s assume that it only produces lactic acid). Well, if I ferment the tea with 10g/L of glucose and take samples at the end of the fermentation (where there is 0g/L of glucose left), I can start drawing hypotheses. If the pH was very high, then the CO₂ must be very high as well (more glucose went into CO₂ production by yeast rather than into acid production by bacteria) etc… the key goal is to give a unique “ID” card to each microorganism, to see their behavior during the experiment in terms of activity (more activity of “x” microorganism means more flavor development of “x” compound).  The ultimate goal is to develop a system through the Pioreactors that could be potentially implemented in other laboratories, breweries, and so on, and to predict and model fermentation.

However, the campaign still has 28 days to go, and any additional funds that are donated (hint!) will be used for system upgrades.

Podcast

Curro and I discuss his research.

The post Modeling Fermentation Through Open Science, with Curro Polo appeared first on 'Booch News.