In mobile telecoms, a charging decision that arrives late is functionally identical to a wrong decision. When a subscriber initiates a call, data session, or SMS, the network node needs a credit decision in milliseconds. Any delay risks SLA breach. Any failure risks fraudulent usage, double-spend, or billing errors that are costly to detect and nearly impossible to reverse cleanly. For i2i Systems, a Turkish systems integrator serving major operators including Turkcell, Türk Telekom, and Vodafone Turkey, the legacy Oracle/TimesTen stack could no longer meet this bar. It was expensive to scale, brittle under peak load, and unable to support the virtualised infrastructure its operator customers were moving toward.
This case study details how i2i Systems replaced Oracle/TimesTen with Volt Active Data as the decisioning core of its Frontiers-CBS Online Charging System, and what changed as a result. It covers the four structural challenges that drove the migration, the architectural decisions behind the Volt deployment, and the production outcomes across latency, availability, revenue protection, and scale economics.
The solution centers on a single architectural principle: move the charging logic to the data, not the data to the logic. All subscriber state lives in Volt’s in-memory partitioned tables. Every charging decision, including credit validation, authorisation, deduction, and policy enforcement, executes in a single ACID transaction at sub-10ms latency. Concurrent deductions across shared household and MVNO bundles are serialised per partition, eliminating race conditions and overspend risk entirely. Active-Active clustering provides 99.999% availability with sub-millisecond failover.
For telecom architects and platform engineers evaluating online charging infrastructure, this case study shows what it looks like when charging correctness, speed, and availability become architectural guarantees rather than operational targets. If your OCS relies on delayed deduction, reconciliation, or scale-up hardware to manage peak load, read on to see how i2i Systems solved the same problems in production.