Construction, Occupancy, Protection and Exposure
The Collection of COPE (Construction, Occupancy, Protection and Exposure) data is essential for organizations to understand an organization’s risk profile and exposure. Without COPE data, underwriters may need to assume the worst, which could lead to higher insurance premiums for organizations.
Considering huge number of variables involved in COPE, we have developed our own feature selection process using statistical techniques to refine list of consumable variables.
After running complex pre-processing techniques over consumable variables, we execute our own custom weighted multi-model approach, which can enhance our model scoring accuracy. This framework is executed on Spark and HDFS to scale as needed. We provide pluggable interface through REST API for easy client on-boarding.