I would recommend DataGalaxy for three main reasons. The simplicity of use, the adaptability to different use cases and the efficient and available support teams.
Overview
A banking group offering financial services to businesses, institutions and individuals, the Crédit Agricole Group has the largest network of cooperative and mutual banks in the world. It employs 145,000 people serving 53 million customers worldwide. Created in 2016 and employing 50 people, Crédit Agricole’s DataLab Groupe’s mission is to develop industrial artificial intelligence solutions for internal use, to support the group’s entities in the creation of Data & AI solutions and to acculturate the Crédit Agricole group to these technologies in an accelerated evolutionary phase.Project background
Originally, the aim of the DataLab Groupe was to explore the potential of data and AI for Crédit Agricole’s businesses and to conduct tests to assess their feasibility. Since 2018, several transformations have been carried out in order to become a reference hub in the creation of innovative, natively industrial, trustworthy and responsible internal data and AI solutions in order to better convince people of the relevance of data and AI.
Their working methods have therefore evolved, to place themselves in a global “trusted and responsible AI” approach, proven on numerous projects carried out in conjunction with Crédit Agricole Group entities. This work took concrete form in early 2022 with the certification and labeling of processes for building, deploying and maintaining its solutions in operational condition.
The DataLab Groupe handles a variety of AI use cases based on different types of data, including:
- Analytical AI: using structured or semi-structured data to anticipate customer difficulties, propose relevant offers or detect fraud attempts or cyberattacks.
- Document AI: using unstructured data to facilitate the recognition and control of documents used in all banking processes.
- Textual AI: exploiting unstructured data to classify e-mails, extract intentions from reports, recognize emotions in text, and so on.
As all AI systems are data-driven, mastering AI systems requires mastering data.
The Data Knowledge Catalog implementation process
The DataLab Groupe needed a tool to centralize the data from all the projects it was working on, including the various data sources from different group entities or from open data, the links between them, the transformations applied, and the links with the use cases developed.
The DataLab Groupe had a few main criteria when choosing a data catalog to meet its needs: to be able to respond to all use cases, to process structured data (banking data present in the database), semi-structured data (navigation logs), but also unstructured data (documents and texts). They chose DataGalaxy because of its flexibility.
The DataLab Groupe began work on its data catalog with a pilot phase involving the documentation of three projects on its main families of use cases: analytical AI, documentary AI, and AI for extracting information from text.
Then, for each of these types of use case, models were defined, structuring the elements of the dictionary and the processing in DataGalaxy to respond to the specificities of each of these types of use case. Today, the application of these models has resulted in greater consistency and efficiency.
Results of using DataGalaxy
Creation of the Open Data Mart,
Internal documentation and communication
A single source of truth
The benefits of DataGalaxy according to DataLab Groupe
- A comprehensive, easy-to-use solution for organizing all the data used by the DataLab Groupe in its various projects.
- The DataGalaxy teams are very responsive, always available with a solution for every question that has arisen.
We’re in the phase of rolling out the tool to all our projects at Groupe DataLab. We run a lot of AI projects that follow the method of labeled and certified projects, which now includes project documentation in the DataGalaxy tool.