AG2R : projects and Machine Learning
AG2R La Mondiale is a French organization offering a wide range of insurance products to 15 million individual and 500,000 corporate clients.
The group wanted to develop new projects based on Machine Learning , but faced several challenges:
Strong Regulatory Constraints:
Scoring Action Imperative
Strong regulatory pressure on insurers to combat money laundering and terrorist financing, forcing industry players to invest heavily in scoring models using supervised learning technologies.
Transition to Production:
Challenges for ML Initiatives
Several Machine Learning initiatives had been launched internally using a mainly build-based approach, but these struggled to scale up and did not reach the production phases.
Absent Data Organization :
Lack of Tools and Methods
The company often lacked the technological tools and methods to operate them necessary to create, monitor and maintain these models.
The DataOps Platform was chosen by AG2R La Mondiale to develop and supervise its scoring models.
Technical Foundation for ML Initiatives
Created a score factory placing Saagie as the technical foundation on which the group’s Machine Learning initiatives are built, starting with regulatory compliance use cases and extending to marketing needs.
Enhanced MLOps culture:
The implementation of MLOps tools and methodologies to automate the supervision of models in production and considerably accelerate the start-up of new projects, from an average of 2 weeks to a few minutes with Saagie.
New MLOps functionalities
The adoption of an approach based on co-innovation between AG2R La Mondiale and Saagie’s Professional Services teams to develop new MLOps functionalities to be integrated within Saagie.
You want to go into production and launch your data projects?