Building and maintaining a customer experience management (CEM) platform

The architecture was explicitly designed to be highly available and fault-tolerant, ensuring 24/7 regime support. It was built as a robust, reliable framework focused on signaling network data. This design supports large-scale processing (3TB/day) by enabling massive aggregation (30B events/day) and the calculation of complex, customizable reports, while also facilitating transparent monitoring and future 5G migration.

Software Mind built the CEM platform using a robust big data stack. Key technologies included Hadoop, Apache Hive, and Spark SQL for processing massive datasets, which were managed by Ambari and coordinated by ZooKeeper. Java and Python were used for development. Ansible, Jenkins, and Git handled automation and CI/CD, while security was managed with Ranger KMS.

The client's main objectives were to build a long-term (10+ years) CEM platform using signaling network data and to create a framework for complex, customizable reporting for internal business units. Critically, they required a highly available, fault-tolerant solution to maintain a 24/7 regime. The project also needed to support future system upgrades and the planned migration to 5G signaling.

Organizations managing massive, continuous data streams would benefit most. This includes other large telecoms, major financial institutions monitoring transactions, large-scale e-commerce platforms analyzing clickstreams, and utility companies managing smart grids. Any industry reliant on 24/7 operations, complex reporting, and processing high-volume sensor or event data (like IoT and logistics) would find it valuable.