Data lab': transforming data into value
At BNP Paribas Cardif, we transform data into customer value and apply our expertise to the data that we collect, enhancing it for the benefit of our insurance customers.
To ensure that we seize all the opportunities offered by data, we have developed a three-pronged strategy: data quality, data protection and data exploitation. At the heart of this strategy, the data lab' is our centre of expertise dedicated to building, operating and reusing our algorithms.
"Data science is revolutionizing our processes and the value we offer to our insurance customers."
The Data Lab is built on three pillars
A powerful IT environment
This powerful IT environment enables us to leverage the full potential of artificial intelligence. For example, we can run advanced models of deep neural networks (Deep Learning) on Graphical Processor Units (GPUs).
Domino: an international platform
Governance of the models we operate from our international-scale Domino industrial data science management platform. This platform enables us to make our algorithms operational by easily integrating them into our information systems and monitoring them.
An open research and development ecosystem
As a member of a consortium launched in 2018 we are contributing financial support to Scikit-Learn, an “open source” software library for machine learning published at the initiative of teams of engineers from the INRIA (the French research institute for digital sciences).
Our analytics managers and data scientists form an acknowledged centre of expertise. Their mission is to enable the data to express its full potential, strictly respecting criteria of ethics and confidentiality, to benefit our partners and insurance customers.
- 4 winners of the Quora international data science challenge in 2017
- 80 data scientists and analytics managers in 35 countries by 2020
- Partnership with the École Polytechnique for training our analytics managers
With algorithms, our analytics manager and data scientists are improving the entire customer experience:
- Customization of offers of products and services
- Pricing as accurately as possible insurance customers’ needs
- Optimised management of contracts
- Accelerated claims management and automated analysis of supporting documents
- 80% of our processes will be automated by 2022
- 75% reduction in supporting documents thanks to our pilot processes
To provide our partners and insurance customers with an ever-improved support, we are continually investing in:
- Automatic document processing: this is a considerable challenge because we process 100 million supporting documents every year
- Natural language processing: our programs will be able to automatically direct enquiries received by email, live chat or telephone, thereby enabling us to respond to our customers more quickly.