NEW TECH FOODS

It’s easy to get caught up in the news and activities of the industry’s global giants, but what about the smaller firms pushing boundaries with bold ideas? In this instalment of start-up spotlight – which celebrates lesser-known companies and their innovations – we speak to Dafna Gabbay, co-founder and CEO of Finally Foods, an AI-driven molecular farming start-up.

Can you tell us about the inspiration behind Finally Foods and how the idea of growing dairy proteins in potatoes came about?

The idea came from an unmet need in the alternative protein space: while demand for dairy alternatives is growing, there was no viable method for producing the most essential milk proteins – like casein – at scale, cost-effectively and with full functionality.

Precision fermentation faces limitations with complex proteins, and traditional dairy is resource-intensive and not sustainable over time. Our vision was to use plants as natural bioreactors, and we identified the potato as a powerful, underutilised host for this purpose.

Your field trial is taking place less than a year after launch – what key factors have enabled such rapid progress?

A few things enabled this pace: first, our exclusive license to a proprietary AI platform allowed us to design the casein and optimise the genetic modifications of the potato efficiently. Second, the deep experience of our R&D team – especially in plant science and AI – allowed us to move quickly from lab to field. Lastly, we’ve had strong alignment with our partners from day one, enabling focused execution.

Molecular farming is often seen as a long-term bet due to regulatory hurdles. How are you navigating these challenges, and what do you see as the biggest regulatory roadblocks ahead?

We’re targeting geographies with clearer regulatory pathways for GMO cultivation, including the Canada, LATAM, Australia and finally the US. We also separate the regulatory track for cultivation from that of the extracted protein, which helps streamline planning.

The main challenges ahead involve ensuring food safety approval for novel proteins and navigating varying GMO-related rules across markets. We’re working closely with regulatory experts to design a global strategy from the outset.

Potatoes were chosen as the ‘bioreactors’ for your casein production. What makes them more suitable than other plant-based expression systems?

Potatoes offer several advantages: they’re genetically well-characterised, relatively simple to modify and already used for protein extraction at scale. They offer high biomass yield per hectare allowing for at least two growing seasons per year.

Importantly, we’ve also developed a cost-effective extraction protocol specifically for casein from potatoes, which improves feasibility at scale.

You’re already in discussions with dairy companies – what kind of response have you had from industry players, and what are the next steps towards commercialisation?

The response has been very positive and there is further news that I'll be able to share in a few months time. Traditional dairy companies are actively seeking sustainable alternatives, but they want solutions that don’t compromise on taste, texture or functionality. Our casein delivers on those expectations.

We’ve secured multiple letter of intents and are moving toward joint development agreements to test and tailor our casein for specific dairy applications.

Scaling up is a major challenge for food-tech start-ups. What lessons have you learned so far about bringing molecular farming to a commercial level?

The key is to design for scale from the start. We chose a crop with a scalable extraction process that uses existing infrastructure. Another lesson is that commercialisation requires deep collaboration with customers–early engagement through JDAs is critical to tailor the ingredient and ensure successful adoption.

How does AI enhance your molecular farming approach, and what advantages has it provided in terms of efficiency and scalability?

Our AI platform helps us identify the optimal gene constructs and expression sites in the plant genome, reducing the trial-and-error phase dramatically. It also enables better prediction of protein functionality and expression levels. This shortens R&D cycles and helps us scale faster, as we can iterate and optimise with greater precision and speed.

There’s an ongoing debate around ultra-processed foods and consumer perception of novel food technologies. How do you plan to communicate the benefits of your potato-based casein to consumers?

We emphasise that our casein is molecularly identical to the protein found in milk, with no foreign DNA in the final product. The production process is transparent and avoids synthetic additives. We plan to communicate benefits through our B2B partners by highlighting functionality, sustainability and simplicity – real dairy protein, just made in plants.

What advice would you give to other start-ups looking to break into the cultivated food space, especially those working with novel protein production methods?

We’re still early in our journey, but one lesson so far is the importance of focusing on the real pain points in the industry, not just the technological novelty. Building a team that combines scientific expertise with business and regulatory understanding is also critical.

Most of all, staying close to potential customers from day one helps ensure you're developing something that truly fits into the value chain and solves a real need.

Finally, what’s next for Finally Foods? Where do you see the company in the next five years?

Our next milestones are scaling up casein production and entering product validation with commercial dairy partners.

In five years, we will prove the commercial viability of molecular farming for casein, establish an impressive portfolio of strategic industry collaborations and start expanding into additional proteins and geographies.