Cognitive machine learning for early childhood development: the Gazouyi start-up takes off

The Gazouyi project first came about in a corridor in the Psycholinguistics and Cognitive Science Laboratory (LSCP), renowned for its studies on infant development. Victor D’herbemont, then a final-year student intern in the CoML* (Cognitive Machine Learning) team, met Xuan-Nga Cao, a research engineer in the same team. Like many parents, the latter had had lots of doubts with the arrival of her first child. At the time, she wondered how she could use the results of scientific research to answer these questions and help other parents looking for the right information about their child’s development.

The project’s name, Gazouyi, came naturally. 

The team is developing a learning module composed of two main complementary parts:

• Professionally designed activities, each associated with one of the 300 levels of child development between birth and five years of age. These activities don’t require any equipment and never involve the use of a telephone: the children are not in contact with the screens, the aim being to discover the early childhood process in a safe family environment.
• A recommendation engine proposes a selection of activities that correspond to the child’s individual needs every week.

Victor explains how the application works:

It is based on active teaching strategies and uses the activities already completed by the family (or not) as input. The more time the family spends on the activities, the more we are able to offer them content adapted to their everyday life

Gazouyi, a guide in the adventure of parenthood

Allowing educational continuity, improving confidence and skills through play, preventing the need for lengthy research on online or in books on early childhood... these are the challenges that Gazouyi aims to address. Surrounded by a network of mentors, experts and early childhood professionals (speech therapist, linguist, psychomotor therapist etc.), the start-up has also developed a partnership with the French Association of Ambulatory Paediatrics (AFPA), whose website Mpedia is a point of reference in the field.

From a scientific point of view, Gazouyi aims to provide insight and make key discoveries about how children grow each day and develop, or on how parenting affects a child’s future, topics on which the startup is also working with the Institute of Psychology at the University of Paris and the School Social Work at the University of Montreal.

*Joint CoML team between EHESS, ENS, the CNRS and Inria Paris, at the intersection between cognitive and computer science. 

© Gazouyi