Case Study: British Gas achieves high-throughput, low-latency IoT performance with Hazelcast

Hazelcast IMDG at British Gas

British Gas-backed IoT business Hive needed a high-throughput, scalable in-memory store to support its Hive Active Heating network (200k homes) and millions of device state caches. The company required real-time querying, easy integration with its backend, linear scale-out and no disk persistence — and after finding MongoDB unable to cope with traffic, Hive turned to Hazelcast’s in‑memory data grid (Hazelcast IMDG) as the vendor solution.

Hazelcast delivered an in‑memory data grid deployed as a cache‑as‑a‑service layer, eliminating backend bottlenecks and enabling easy cluster expansion. With Hazelcast, Hive achieved roughly 20,000 writes/sec on a 20‑node cluster with average latency under 1 ms, simplified management and linear capacity growth; Hive has added Hazelcast support and plans to expand nodes as its product line and deployment footprint grow.

British Gas