How AI is Powering Personalised Food Experiences: Insights from Personalised Nutrition Innovation Summit
At this year’s Personalised Nutrition Innovation Summit Europe, Nick Holzherr, CEO and Founder of Whisk.com, presented observations on personalised nutrition and the technology powering personalised food experiences.
“Consumers demand more personalised experiences, better visibility of ingredients, nutrition, and allergens. AI is making it easier for people to shop for their preferred food choices.”
– Nick Holzherr, CEO and Founder of Whisk.com
The Personalised Nutrition Innovation Summit gathers industry leaders, tech disruptors and leading nutritional experts to understand the potential of this growing field on public health. In particular, how the use of data can help multinational food companies, healthcare organisations and retailers to help people eat better and live well through personalised nutrition.
Personalised Nutrition: Today and the Future
A deeper understanding of the user (their preferences, behaviour, biometrics, conditions and context) is essential when it comes to personalised nutrition. Whisk’s personalisation system builds this understanding by collecting a wide range of data from both explicit and implicit user behaviour.
Taking into account diets or allergies, ingredients likes and dislikes, health metrics, family size, or favourite dish, Whisk’s technology offers personalised recipe recommendations based on the data user provides.
If you’re a vegan who dislikes mushrooms and are a confident cook, you can easily modify your preferences to match your recipe recommendations.
Based on user behaviour, our technology can also offer recipes based on viewed, saved, shopped, shared or trending recipes.
Finally, using data around context, Whisk’s technology offers personalisation around food events, weather, user inventory, store deals and popular recipes.
Whisk also allows the user to easily personalise and swap ingredients in their shopping list or cart to match these preferences.
Impact on Health
Whisk’s technology allows us to understand the composition of different foods and recipes to calculate nutritional values for any recipe, including Glycemic Load and Glycemic Index. “Health Score” helps people understand the impact of a recipe on their health. Take a look at showcase.whisk.com.
Although health concerns do not always drive personalised nutrition, finding the perfect recipe, creating bespoke meal plans and getting ingredients is a compelling use case for people with diabetes.
Earlier in 2018, Ascensia awarded Whisk the prize to combine our data & technologies to recommend recipes based on blood sugar responses.
The next step will be to build an AI engine to personalise recipes based on an individual’s blood sugar readings and how they will respond to different food types.
The Importance of Data and Data Normalisation
Increasingly Whisk is working with global businesses who have recipe content spread across multiple markets and languages, stored on legacy systems, with incompatible formats, disparate tags and requiring extensive time and knowledge to update them.
The solution is to use Whisk’s platform to ingest the content, reformat it with micro-formatting, calculate nutritional values and scores and tag with up to 200 different labels from dish type, complexity, cuisine, etc.
Once it’s in the Whisk platform, as new data sources update – such as USDA ingredient databases – every recipe can be automatically updated with no additional effort. New tags and rules can be added for all our clients to use or kept exclusive and all the data be delivered into any platform via API. Once tagging and nutritional data is accurately mapped, personalisation can begin to add real value to users.
“The cost savings can be in the millions but the opportunity and value of driving SEO performance, enhancing user experiences and freeing up nutritionists and tech teams to focus on adding more value elsewhere is far higher.”
– Nick Holzherr, CEO and Founder of Whisk.com.
As the connected home looks to structured recipe content to power kitchen appliances, such as switching on ovens or setting pans to simmer, this structured data is set to become far more critical and the risks of being left behind far great.