Nutrition and tech experts bring carb-coding power to the masses with AI startup

Performance Nutritionist David Dunne has spent a decade supporting some of the biggest names in world sport, including Olympians and world champions.

And it was while undertaking a PHD in nutrition, behaviour change and technology, that Dunne met many other like-minded professionals who were fascinated by the impact that technology can have on behaviour.

He explains: “We started to work on research together for about three years and realised that with our shared expertise we were able to build large scale behavioural interventions that could affect people’s behaviour.​

“There was a London-based VC that heard about the research we had been doing and he said ‘stop wasting your time researching and build a company already’.”​

Taking this advice on board, they joined UCL’s (University College London) startup incubator programme Conception X and started to build ‘Hexis’.

The accomplished team of co-founders includes: Xiaoxi Yan, Duke-NUS Medical School PhD candidate, data scientist Dr. Rodrigo Mazorra, behavioural science expert Dr. Carmen Lefevre-Lewis, academic and exercise physiologist Dr. Sam Impey, data scientist Dorota Jagnesakova, Performance Science Advisor Dr. Brian Cunniffe and marketing advisor Aman Kumra.

Dr Sam Impey, who published a paper in 2018​ that disrupted sports nutrition. He introduced the notion of carb periodisation, later coined 'carb coding' by the Hexis team. This refers to the ability to enhance performance, recovery and adaptation to exercise through the personalised periodisation of carb consumption.

The team spent two years building an algorithm that could provide personalised carb coding advice at scale. The user tells the app their lifestyle information, body measurements, training schedule and goals and the app provides breakfast, lunch, dinner, and snack recommendations which are translated into recipes which meet their specific needs, according to their training demands.

The team has created a portfolio of 1,500 recipes using simple, easy-to-obtain ingredients and the app simplifies the nutrition suggestions into a

traffic light colour coding system – low, medium and high carb.

“At our core we are on a mission to express human potential using technology to translate complex science into simple and personalised nutrition​ practices, to allow people to feel and perform at their best," ​Dunne says.

“Carb coding allows you to manipulate how your body adapts and optimises how we perform and recover. Traditionally this type of nutrition advice has only been available to elite athletes but we have built an algorithm that can deliver this at scale to help more people express their full potential.”​

After creating the algorithm, the team raised £300k, primarily from professional athletes interested in using the service, and with that they built the beta version of the app which was rolled out last year in a 12-week trial involving 1000 athletes.

“That went very well and we got some phenomenal feedback to the extent that some of the athletes have paid to carry on using the beta model.”​

The team currently has a CrowdCube fund raiser​ open to raise money for commercial launch of the product next year.

Dunne says the venture is drawing a lot of interest and investment from professional athletes and sports nutritionists.

“There are lots of personalised nutrition apps on the market now but most of these are based on retrospective information, with the user just counting calories," ​adds Dunne.

“When it comes to optimising performance, you need to know how to distribute your nutrition throughout the day and week according to your training schedule in a predictive way so you can know what you should be eating in order to get the most out of your workouts.​

“Our app will always be based on the most up-to-date scientific evidence as our team stays on top of the latest research in this field.”​