StoneX taps into real-time efficiency and unlocks new client products

Challenge 

Metrics, logs, and stream processing

StoneX is a Fortune 100 financial services franchise that provides global market access, clearing and execution, trading platforms, and more to clients worldwide. With over 330,000 active retail accounts across 180 countries, StoneX generates terabytes of logs per day — a number that’s rapidly growing.

Handling trillions in annual trades demands real-time efficiency and scalability, but with Kafka, StoneX struggled to meet its performance and latency requirements. Not to mention bloated costs that hindered scaling and less-than-helpful tech support from its managed Kafka service provider.

StoneX was searching for a more efficient solution to power its high-throughput observability pipelines and latency-sensitive streaming data processing, while reducing overall costs.

Why Redpanda 

Simple, powerful, and efficient to run

Based on its experience with Kafka, StoneX chose Redpanda for several reasons:

• Significantly lower latencies
• Scalable and cost-efficient
• Friendly and reliable tech support
• Easy migration from Kafka and legacy batch-processing 

During testing, StoneX was impressed by Redpanda’s ability to handle the same workload on far fewer resources than Kafka, translating to generous cost savings. The team was also pleased with Redpanda’s seamless compatibility with the Kafka protocol, which made migration a piece of cake. 

“We really did just change the endpoint, and off we went,” says Michael Pearce, Senior Architect at StoneX.

Results

Real-time efficiency that creates new client products

In trading, every millisecond counts. Since rolling out Redpanda into its production systems, StoneX’s latency dropped from double-digit milliseconds to single-digit milliseconds, significantly optimizing workflows such as trade execution and machine learning prediction model training.

With Redpanda, StoneX can also handle the same workloads on 5-6x fewer resources than Kafka, providing the team with a low-cost and highly scalable logging and metrics architecture. Furthermore, StoneX leveraged Redpanda to modernize its legacy data processing from batch to real-time streaming. This makes data readily available to the business and enables StoneX to create additional client products from a streaming data flow, like financial vehicles and tradable ETFs. “Redpanda is enabling new data products that were impossible with Kafka,” Michael says.

Thanks to its real-time efficiency and single-digit latencies, Redpanda continues to power StoneX’s retail trading platforms for low-latency trades across its global operations.