Open Source in Finance Forum New York 2022

Low-latency, high-performance electronic trading networks predate the emergence of cloud computing by more than a decade. Due to the stringent performance demands of this workload, schemas are explicit, encoding is binary and transports are unreliable. Cloud computing, on the other hand, favors self-describing text-based encoding and stateless transports like HTTP. This results in excellent interoperability between services and teams, but results in inefficiencies that compound at scale. This talk will discuss the challenges of bridging these two worlds and illustrate patterns that the speaker has used to solve the problem of analyzing granular electronic trading data in the cloud. The audience will be taken through a history of common electronic trading protocols such as FIX, ITCH and Simple Binary Encoding. The speaker will identify the differences between high and low-context encoding, and describe the obstacles encountered ingesting common electronic trading formats into cloud-native services. The session concludes with a pattern for integrating the two worlds harmoniously by deconstructing schemas, encodings and transports.