top of page
20180409 DataCity new logo2-09.png


Boost employment in local shops and businesses

How can new restaurateurs be helped to predict their business activity more accurately in order to manage their resources more effectively and ensure the long-term success of their business?


Every day in Paris, more than 3 restaurants open, while a further 6 cease trading. The life expectancy of a restaurant used to be 7 years, but now it is 2 years. For a new restaurateur, this brings increased financial pressure and makes it difficult to obtain a loan, thus reinforcing the need for business predictions.


2 years

is the lifespan of

a restaurant in Paris



Cartes Bancaires CB, the leading payment brand in France (80% of payments received in the restaurant trade) and the City of Paris opted to work with DreamQuark, a startup which was voted FinTech of the year, and specialises in developing platforms that enable financial services to use a deep learning model. Together, they have developed an activity prediction model for shops and business to help them to anticipate their activity from the very first month of trading. The aim of this API is to offer a decision support tool to new restaurateurs to help them manage their finances more effectively and thus grow their businesses. During the two-month experimentation period, they cross-referenced CB card payment data with external data affecting business activity and developed the first model to predict activity peaks for restaurants with an 80%+ accuracy rate.



DreamQuark is developing a platform named "Brain" to promote the mainstream use of artificial intelligence. Based on cutting-edge Deep Learning technology, Brain offers its bank and insurance clients unparalleled interpretable results. Brain covers all of their areas of activities with applications dedicated to fraud, risk, marketing, and compliance, including policy underwriting, asset management, compliance and fraud, and also customer segmentation and retention.





The DreamQuark startup worked with CB to develop a decision support tool based on the predictive model of CB’s exclusive anonymised data, enhanced with external data affecting shops and businesses in the district. The Irma solution allows traders to anticipate their business activity using a dashboard showing payment fluctuations, reservations and resources allocated. Once fully developed, the API will be easy to integrate into the trader’s business environment.



bottom of page