Societe Generale set to use data and AI to help it transition and success to the digital space to give SocGen an edge in personalised customer services, responsiveness and speed, efficiency and risk controls.
As every organization or company will, the 150+ year old French lender faces up to a host of challenges and challengers hence, has built an army of more than 1000 data experts, including 65 chief data officers, working across all of its businesses and geographies.
The bank has more than 330 data use cases in production, of which nearly 170 are based on Artificial Intelligence (AI), covering areas such as start-up applications, facial and biometric recognition, automatic credit ratings and analysis tools for market activities.
Frédéric Oudéa, CEO, Societe Generale, says: “The increased and responsible use of data, notably with AI, is a strategic asset on which Societe Generale is continuing to capitalise to adapt and improve its existing distribution and production models, but also to invent and develop new alternative business models. Becoming a data-driven bank, with a responsible use of data, is at the heart of our strategic agenda.”
To harness the full potential of data, the bank has developed a main hybrid datalake platform combining private and public clouds.
Societe Generale has also set up tools to ensure data quality, facilitate data access and sharing in accordance with regulations, and industrialise and generalise uses. Some 80% of servers are now hosted in the first and second generation clouds.
By 2025, the bank has set the objective of using 75% “second generation” cloud. including 50% private cloud and 25% public cloud, to allow developers to access off-the-shelf services such as machine learning modules and databases.
Meanwhile, a “cultural revolution” is underway to ensure staff are ready for the data-led business: management is going through a programme to foster knowledge about AI, while staff around the world are being offered the chance to get training and also to join a re-skilling programme.