Brains and AI: Shaping the Future of Food  

How did your career path lead to the founding of Thimus? 
In practical terms, very casually. I started my career as an innovative entrepreneur in 2002, in a completely unrelated industry (metal finishing). We co-patented a very innovative diagnostic system for hot zinc plants and I ended up licensing it around the world and opening 4 offices in 2 continents. In the end, in 2014, we sold the company I had co-founded to a larger group and I left after two years. I was looking for a new opportunity and somebody at a dinner mentioned applied neuroscience to me. I came as a curious investo and stayed for the passion and the sense of purpose I did get after the pandemic. I wanted to have an impact on food systems and Thimus can definitely be an opportunity to do so.

Which nutrition projects is Thimus currently working on? 
We are working on several food-related projects, some of which have very interesting nuances in nutrition. For example, we have helped developing with Alamance Food Labs in North Carolina and Savor a completely revolutionary whipped cream that uses a fat component made from CO2. It’s an amazing ingredient and it’s carbon neutral. Also, last month we developed in a rapid iteration a ready-to-drink cold coffee brew, called NeuroBrew, that contains functional ingredients coming from asian traditional medicine and a milk component that is plant-based, protein rich and very low in sugar. Both these products are meant to showcase how sensory neuroscience can become the key to consumer loved products that are sustainable for the planet and for human health. 
Now we are about to announce e very big project on craving and liking in GLP-1 agonists consumers. But that is still confidential for a short while!

Could you illustrate with a practical example how Thimus leverages AI for its projects? 
I believe the simples example is that we use AI to recognise brain signal that is meaningful for emotional and cognitive processes, eliminating the “noise” coming from chewing and other muscular activities. But a deeper example is possibly the use of AI to unearth insights on the correlation between brain response to food consumption and demographics, habits or even specific ingredients or other technical features in food and beverage formulation. AI can then illustrate these insights and deduce suggested action in our platform, guiding product developers. 

How are the AI-driven results of your measurements incorporated into new food products? 
The AI model we are developing actually has an impact on new food products in two main manners. Firstly, our AI model analyses both the neuroscience datasets from the prototype testing and the declarative data we collected, presenting users with extremely friendly ways of interrogating outcomes and getting actionable insights on formulation or other tested aspects. Furthermore, the model learns from all of our global database of ever-growing data, which allows to infer additional insights from simila demographics, or similar products and product categories. So AI maximises the strong value of brain data by compounding it with a lot of other relevant datasets. 
Secondly, we just started a special project: private instances of our proprietary AI will be made available to selected partners. In this particular scenario, not only the user will have access to AI-driven insights based on Thimus data, but they will also be in a position to actually drag and drop their own datasets (i.e.: sales data, formulation data, etc) in the model and train it to produce even more advanced, well interconnected insights to enrich their foresight. 

How might the use of AI transform the nutrition industry? 
In many directions, for sure. To me, that doesn’t mean all these directions will automatically be good. We need to pay attention to what are the training rules for the AI we want to deploy. That’s why I’m happy we are basing our model on extensive emotional and cognitive data from real humans: humans are the anchoring point, otherwise synthetic data leads to synthetic products, stuff that is not made for humans but only for convenience or margins. That doesn’t work for us. 
In reality, I would expect AI to allow a much more targeted level of nutritional offer, based on both demographics and cultural profiles. AI could help local bioregional food systems improving their product offering, making them more creative to promote locally made, or grown, products, to enhance that de-centralization movement in food systems that would make a lot of sense. 

What are your personal ambitions? 
 I confess I am still, at 52, a big dreamer. My personal ambition with Thimus is to leave a mark on the food industry and space in general. I’d love to know that my efforts and those of my team can bring to life a new tool, a new perspective, new data and insights to guide into a new era in food products and experience design. I would love to know that I could have been the promoter of a humanistic wave in F&B!