Predict paiement variations to facilitate HR, stock and reservations management for the week ahead
Prediction of HR needs for small businesses
// KEY RESULTS
The decision making tool helps shopkeepers to anticipate their resources management, based on peak prediction at local scale. After only 1 month of activity record, the solution will provide : paiement and reservation expectations, plus the resources variations planned accordingly. This predictive model is based on anonymized data coming from local transactions, increased with external data related to contextual element affecting shoppers behaviors
The experimentation took place in Paris.
// PARTNERS INVOLVED
Partners are providing the following data sets: weather, events calendar, calendar of cultural activities, local community outreach programs, scraping data from social media and third-party professional websites. The selected startup – DreamQuark - offers AI solutions for Financial Services through a platform that automates Deep Learning model creation.
// PAIN POINT & OPPORTUNITY
The restaurant sector, for example, is a seasonal industry affected by multiple types of external events (80% of annual turnover in two months). Because restaurants tend to evaluate their own needs, incorrect estimations of HR resources can have a major and direct impact on revenue.
Every day in cities like Paris, more than 3 restaurants are created while 6 others close. Before, the life of a restaurant was 7 years, today it is 2 years. For a new restaurant owner, this means increased financial pressure, making access to the bank loan more difficult and reinforcing the need to predict business activity.