MAPCITE Scales Out With MongoDB

Insurance and Banking industries create and need enormous volumes of data to support their risk management operations. MAPCITE's Location Analytics solutions make it possible for our users to blend together many different types of data to create answers to extremely difficult questions. For example, in mortgage risk, combining 78 different pieces of data that pertain to a particular building, answering the question “should we lend against it?”. Imagine the size of the system needed to make available these 78 datasets, against every property in the world, and that’s just where we get started. Now imagine every insured location and the risk models that govern how insurance companies price their risk and manage their exposure. We are talking about very big data that needs immediate accessibility.

MAPCITE's Location Analytics are computed using NVIDIA CUDA GPU processing and this is directly connected to MongoDB that delivers a massive horizontal scale, highly secure and reliable back-end data store.

Screen Shot 2016-08-20 At 14.14.44

Millions of insured assets affected by the NYC storm surge, one click calculates exposure in real time.

Why MongoDB?

We needed a powerful NoSQL solution to help manage the ever increasing volumes of location data without sacrificing core application performance. We knew Mongo could scale horizontally but with the later releases introducing 2dShpere spatial indexing and querying, we had found a great fit for our needs.

Being able to route queries efficiently through huge amounts of data and pull only the data we need, means we can give our users a great experience irrespective of data volume.