Case Study: Leading Global Online Payments Company achieves real‑time fraud detection and multimillion‑dollar savings with Software AG (BigMemory)

Better, faster fraud detection drives higher profits

A leading global online payments company was losing money to fraud and struggling with slow, disk‑bound rule processing. Its legacy platform could only handle 50 fraud rules even though adding rules (one alone would have saved $12M annually) and scaling to thousands was necessary to reduce fraud (about $0.30 per $100) and meet stricter latency requirements — the existing 800 ms SLA was missed about 10% of the time, costing roughly $10M a year.

The company deployed BigMemory Max as a TB‑scale in‑memory rule store, increasing capacity to thousands of rules and raising throughput from ~2,000 to ~65,000 transactions per second with predictable low latency. The BigMemory solution (currently a 4TB store) enabled the 650 ms end‑to‑end SLA for about 99% of transactions, delivered an estimated tens of millions of dollars in annual profit from reduced fraud losses, and saved roughly $1M a year in avoided database licenses; the firm plans to expand to 150TB and target a 250 ms SLA.