MOA Foodtech Is Brewing Egg Replacers From Factory Waste, and an AI Picks the Microbes

On October 14, Chemical and Engineering News put a small Spanish company on its annual list of 10 Start-Ups to Watch. The company sells no food to consumers. It sells the ingredients that let food companies cook without eggs, without synthetic emulsifiers, and increasingly without animal inputs. Its raw material is cheaper than that sounds: the starch-rich leftovers that other food factories throw away.

MOA Foodtech, based in Noáin in Spain's Navarra region, is in the matchmaking business. On one side, a library of more than 300 microbial strains. On the other, agrifood side streams that currently cost their producers money to dispose of. Between them sits Albatros, the company's in-house AI platform, which predicts which microbe will thrive on which waste stream and what fermentation conditions will turn that pairing into a sellable ingredient. In early October, the company formalized its most significant commercial test yet: a Bakery Innovation Alliance with Espiga I+D, a cereal research centre near Pamplona, to validate its egg replacer on real industrial bakery lines.

An AI scout for fermentation

Traditional ingredient R&D is trial and error: a scientist tries a microbe on a substrate, adjusts temperature and pH, waits, measures, repeats. MOA's claim is that predictive modeling can skip most. According to the company, Albatros evaluates fermentation scenarios computationally and recommends the most promising biological process for a given byproduct, cutting development time for a new ingredient from six months to under two weeks. That figure is company-reported and has not been independently verified, but it captures the bet: that bioprocess discovery is becoming a data problem.

The microbes require constant care. With 300 strains to keep alive, MOA's team of biochemists and bioinformaticians works late into the night, and chief scientific officer Susana Sánchez-Gómez told C&EN the staff have their favorites. All of the company's strains are non-GMO yeasts with established safety status for food use in the United States and Europe, which matters commercially: the ingredients can be declared simply as "yeast" on labels and, under current European rules, do not require Novel Food authorisation.

Two ingredients, one pitch

The flagship is MOA Q5, introduced in late 2025 as a single ingredient that replaces egg and functions as an emulsifier. The company reports that it replicates egg's emulsification, foaming and coagulation across bakery, pastry, pasta, plant-based desserts and sauces, while maintaining volume, structure and bite. It is also positioned as a substitute for E-471, the mono- and diglyceride emulsifier common in industrial baked goods that clean-label reformulators are trying to remove. One constraint travels with it: Q5 contains gluten, so allergen labelling applies.

The second ingredient, MOA Yeast, is a nutritional and functional booster aimed at plant-based proteins. The company describes it as improving taste, texture and stability while reducing sodium and masking the off-notes of plant proteins, and says the biomass contains all essential amino acids with a protein digestibility score of 0.9, which it presents as on par with soy, beef, eggs and casein. A pet-food variant, MOA Yeast FeG, is already used in a commercial pet food product, though MOA would not name the customer. Both ingredients are sold to food manufacturers rather than consumers, through MOA's in-house sales team, and manufactured by a contract development and manufacturing partner in Spain.

Food factories pay to get rid of the feedstock. MOA's proposition is that an AI can find the microbe that turns it into an ingredient, and a bioreactor can do the rest.

Why bakeries are listening now

The Bakery Innovation Alliance arrives at a moment of genuine pain for industrial bakers. Egg prices have been unstable for three consecutive years, with repeated avian flu outbreaks making supply difficult to forecast. At the same time, cleaner-label demand is pushing manufacturers to review additives like E-471. Finding replacements that perform reliably outside controlled conditions has been the hard part.

That is the gap Espiga I+D is built to close. The centre, operating since mid-2025 in Sarriguren, Navarra, runs from laboratory to semi-industrial production, including a Bühler teaching mill capable of processing more than 14 tonnes of flour per day and industrial bakery lines with lamination and rounding technology. Under the alliance, manufacturers can test ingredient samples and pilot formulations on real equipment before transferring recipes to a factory. "The bakery and pastry industry faces more pressure than ever to reduce its dependence on egg," MOA chief executive Bosco Emparanza said when the partnership was announced.

The money and the scale-up

MOA's funding history is a map of European agrifoodtech support. The company took a 100,000 euro participatory loan from the regional agency Sodena in 2021, then raised 1.5 million euros in 2024 in a round led by ICOS Capital to launch its products and build the Albatros platform. In September 2026 it closed a 3.3 million euro round co-led by the European Innovation Council Fund, Swanlaab Innvierte Agri Food Tech and CDTI-Innvierte, with participation from Sodena, Viscofan and Clave Capital. Total funding now stands at 12 million euros, roughly 14 million dollars.

The September capital is earmarked for scale: MOA plans to grow fermentation capacity from 120,000 to 225,000 litres and expand distribution of its first three ingredients across human food, pet food and, increasingly, nutraceutical applications. The company says postbiotic and mental-wellbeing ingredients are in development. It also runs co-development projects with industrial partners that want their own side streams converted into functional ingredients.

What remains unproven

MOA is still a small company selling into an industry that rewards reliability above novelty, and its story has caveats. The headline figures, the six-months-to-two-weeks development claim, the 0.9 digestibility score, the egg-equivalence of Q5, are all company-reported; no independent third-party validation has appeared in the coverage so far. The gluten in Q5 limits its use in allergen-free formulations. The pet food customer is confidential, which makes the "already commercial" claim hard to audit. And at 225,000 planned litres, MOA's capacity is a rounding error next to the fermentation giants, so growth depends on convincing risk-averse manufacturers to reformulate established products around a startup's ingredient.

It is also worth distinguishing what MOA does from precision fermentation, the better-known food tech approach explained here. Precision fermentation programs microbes to excrete one specific recombinant protein, such as dairy casein. MOA practices biomass fermentation: the whole yeast cell is the product, grown on waste, with AI choosing the strain and the recipe. It is less glamorous and less patent-thick, but it sidesteps the costliest parts of the precision fermentation playbook: engineered strains and expensive purified feedstocks.

Whether that trade-off wins will be decided in the bakery trials now underway. If Q5 survives lamination lines and 14-tonne flour days without degrading texture or volume, MOA will have something most food tech startups lack: proof that its ingredient works where food is actually made, not just where it is invented. For an industry trying to do more with what it already throws away, that is the metric that matters.