Alpaca unlocks 100x faster order processing for high-volume algorithmic trading

Challenges

Performance and scale without the growing pains

Alpaca offers modern financial services APIs for crypto and stock investing, real-time market data, and end-to-end brokerage infrastructure. Its custom-built Order Management System (OMS) is designed for ultra-low latency and scale to accommodate millions of investors worldwide.

After years of high-volume algorithmic trading, Alpaca set out to introduce major performance improvements into its OMS. The OMS is the “middle man” between every client order and the market. And, as the backbone of Alpaca’s trading service, the OMS needs to process orders efficiently.

To roll out a faster and easily scalable version of their OMS, Alpaca needed a high-performance streaming data system that could keep up with the company’s rapid growth without introducing complexities.

Why Redpanda

Simplicity and performance all in one package

Initially, Alpaca considered Apache Kafka^® for its familiarity, but hesitated due to the administrative overhead and configuration complexity often required in a production environment. Instead, they chose Redpanda for several reasons:

• Simple to deploy, run, and manage
• Impressive performance with minimal configuration
• No need for JVM or other dependencies
• Developer-first tooling that reduces friction

“Redpanda stood out by having a single binary we could deploy and run, and have an entire streaming engine ready to go. The performance, throughput and latency we got with almost no configuration was incredible,” says Raja Bhatia, VP Engineering at Alpaca.

Alpaca was also pleased by Redpanda’s generous support from development to production. For instance, while building a write-ahead log — a standard method for ensuring data integrity — Redpanda introduced Alpaca to the creator of franz-go, the Go client that interacts with Kafka within the OMS. With his help, they tailored franz-go to their use case, enabling enhanced data durability and recovery.

Results

Better performance, simpler operations, and stellar support

With the help of Redpanda’s lean streaming data platform, Alpaca rolled out a new and improved version of its OMS that delivers 100x faster order processing and consistently low latency — even during heavy trading volumes.

Thanks to Redpanda’s straightforward infrastructure and smooth developer experience, Alpaca reduced its order processing time from ~150 milliseconds to consistently under 1.5 milliseconds. Since order processing is sequential per account, optimizing order processing time gets the order to market dramatically faster, providing a better experience for Alpaca’s customers.

“We highly recommend checking out Redpanda’s streaming data platform for its speed and simplified deployment,” Raja says.

While Alpaca is thrilled with its OMS's significant performance improvements and seamless scalability, the work continued with further improvements to reduce latencies and enhance data throughput. With Redpanda’s ongoing support, Alpaca plans to leverage Redpanda’s powerful streaming data capabilities across many more services.