Case Study: FICO achieves sub‑1ms fraud detection and 24x365 uptime with Couchbase

A Couchbase Case Study

Preview of the FICO Case Study

Supporting the world’s #1 fraud detection platform with Couchbase

FICO, the world’s leading fraud detection provider that scores about 65% of the world’s credit and debit cards, needed a highly available, ultra-low-latency datastore to support real-time credit checks, fraud screening and targeted offers for new telecommunications customers both in-store and online. Relational systems couldn’t scale to the required throughput and existing HA/DR solutions required heavy custom development, so FICO evaluated NoSQL options and selected Couchbase for its speed, scalability, cross-datacenter replication and ability to persist large XML objects.

Using Couchbase Server as a Customer 360 platform and user profile/caching store (leveraging N1QL, multi-dimensional scaling, cross-datacenter replication and an in-memory architecture), FICO achieved sub‑1ms response times, a complete HA/DR solution delivering 24x365 uptime, and direct key-value access for neural-network fraud algorithms — meeting its performance and availability requirements at scale.


Open case study document...

FICO

Claus Moldt

Chief Information Officer


Couchbase

55 Case Studies