The bank, fertile ground for low code no code in data science

Dataiku’s slogan: everyday AI, or the democratization of artificial intelligence. The platform is thus used by data scientists to industrialize projects. This is for example the case of Sephora, whose desire was to accelerate the life cycle of AI models.

These concerns are not sectoral, however. The finance sector, and in particular banks, is also expressing growing interest in AI solutions, in particular to equip citizen data scientists with them.

Data scientists refocused on their fundamentals

Within Crédit Agricole Normandie-Seine, the Dataiku application is intended for two populations, as explained by Pierre Pilet, data science manager within the regional bank.

The data science team develops valuable projects around data and market knowledge. The purpose, however, is to embed the ability to scale up in the data products.

In this perspective, CATS, the group IT department of the 39 regional banks of Crédit Agricole, has initiated a proof of value of the Dataiku platform. The opportunity for Pierre Pilet to refocus the role of the data scientist on its fundamentals, namely the analysis and development of uses.

However, until now, data scientists had to turn to Hadoop platforms (called Zeus). The disadvantage: the obligation for these experts to devote themselves to tasks devolved in principle to data architects and data engineers.

Dataiku must therefore allow a refocusing of the role of the data scientist, the infrastructure and architecture aspects being entrusted to the publisher and to CATS. For Pierre Pilet, the democratization of these solutions must contribute to bringing data scientists and businesses closer together.

The ambition of self-care in data science

Initially, data collaboration will focus on projects related to, for example, scoring and segmentation. The data science manager anticipates a second step: the development of self-care in the use of data.

It is thus a question of allowing collaborators of studies, credit or marketing, for example, to gain in autonomy by seizing the data analysis themselves to cover a first level of need.

This so-called self-care approach is also pursued by the new entity SG Retail France, a merger since 1^er January 2023 of the Societe Generale and Crédit du Nord networks. Thanks in particular to low code / no code, its Data & IA director, Karim Perdreau intends to increase automation.

The objective is also to avoid transforming the data department into a bottleneck, which requires the introduction of a part of decentralization. The ambition: to provide off-the-shelf or self-service tools to business lines.

Other banking players also use technologies such as Dataiku to meet the needs of data scientists and citizen data scientists. This is the case, for example, of La Banque Postale, which relies on a hybrid data organization, with data skills centrally and in certain businesses.