AI-FOOD: AI-based mobile app

AI-FOOD: AI-based mobile app with recommender system for food retailers and clients

One of the main challenges faced by today’s over-informed population is to rely on trusty sources to receive information and advice about how to improve their food habits and choices. People are more and more aware about the need to shift to healthier and more sustainable habits, but many times, the information is not clear or they rely in not contrasted sources as social media.  Having the whole food system involved in the goal of going towards a more transparent and better food future is a must in order to achieve a healthier population and to reduce the environmental impact of feeding to all of us.
Supermarket chains can make an impact through offering a novel approach to support consumers in understanding their nutritional needs and habits and in identifying those products that best meet their requirements.

The aim of this project is to put in the market a solution that will help food retailers to contribute in the path to a better food system by two ways: in one hand, with a recommender system that gives consumers personalized information and advice about how to improve their food habits and choices. In the other hand, helping to the chains enhancing the engagement of their consumers in a way that the brand becomes the first choice when doing their purchases. 

We will use two features of the consumer platform T-Assist (https://t-assist.eu/, tha... was created to gather very valuable and trusty information on how good or bad the nutritional habits of families with children are.  Other goal of the platform was to offer a variety of tools from the field of market-analysis to gain insights into the consumer behaviour and preferences of families with children living in Europe. These tools provide access to high value data and analytical services, thereby providing an opportunity to generate revenue and develop strategic partnerships with stakeholders in the food sector.

The features of interests are:

• a reporting system of the purchases done in different supermarkets. It will be used to develop an app using picture-based machine learning techniques to get information from grocery receipts. This will allow to know what kind of products consumers are buying from the competitors, to overcome the bias inherent to self-reported data and to know more about the consumer habits and preferences so we can try to engage them with different strategies.
• an AI-based recommendation system to improve nutritional habits of families with children. It will be fitted to Eroski’s needs and will take into account the information of the loyalty card from a test group from the Eroski program Friends & Family.

From the previous work we will develop an app integrating an AI-based recommender system and able to collect the information of purchases done taking pictures of grocery receipts, conceived to be exploited by supermarkets and retail chains that have loyalty card programs implemented.  With the data collected from loyalty card and the app, supermarket chains could bring tailor-made recommendations to their consumers with different purposes (promoting healthier choices, offers on substitute products of the competition, on local products, on short expiration date products, …).
This way, it could improve the information exchange between consumers and the brands.

By allowing data collection and analysis, the app could also help to the brands predicting consumer behaviour patterns such as shopping frequency, format sizes, etc. avoiding stock shortages and other supply issues.  And last but not least, giving to their consumers personalized recommendations of their interest such as offers or healthier suggestions, and help to focus education campaigns towards healthy choices.