BNP Paribas Launches LLM as a Service for Generative AI

BNP Paribas has introduced an internal LLM as a Service platform to accelerate the deployment of generative AI use cases across its businesses, announced in a press release. This platform provides the bank's entities with access to a variety of language models through a secure and standardized interface.

The platform is hosted within BNP Paribas' data centers and is equipped with specialized computing capabilities, such as GPUs. It offers access to open-source models, models from Mistral AI, and soon, models trained on internal datasets tailored to specific needs. This setup allows for efficient execution of language models while managing costs and ensuring data privacy compliance.

Several generative AI applications, including internal assistants and document generation, are already in production or testing phases within the bank's various entities, such as Hello bank!. The platform's deployment is part of a broader strategy to integrate AI into existing processes, enhancing operational efficiency and innovation across the group.