Software Engineer

C++ Engineer - High-performance Trading Systems

Company Insight:
A leading global proprietary trading firm is expanding its US cash equities business. This team operates at the frontier of systems engineering - building infrastructure that processes market events in microseconds, sustains hundreds of millions of operations per second, and scales to handle the most demanding workloads in electronic markets. If you've built the data and feature infrastructure that feeds production AI systems at scale, the translation to trading systems is natural - and the performance bar is one you'll recognise immediately.

Responsibilities:
Design and build the core execution stack: order and execution management systems that ingest and respond to live market data from major US exchanges in microseconds or less.
Architect high-throughput key-value stores and data pipelines supporting read/write loads in the hundreds of millions of QPS - and push those numbers further through architectural and systems-level optimisation.
Own the performance surface end-to-end: memory layout, cache behaviour, data serialisation, storage efficiency, and access cost - the same disciplines that matter in large-scale ML feature infrastructure.
Collaborate with FPGA engineers to integrate hardware acceleration into a modern C++ software stack.
Work alongside quant researchers and traders to improve pricing models, execution logic, and decision-making tools.

Requirements:
Strong C++ engineering - you write production-grade, performance-critical systems and reason clearly about what the hardware actually does with your code.
Experience building and operating petabyte-scale data infrastructure, high-QPS storage systems, or training pipelines where latency and throughput are first-class constraints.
Solid CS fundamentals: data structures, concurrency, storage systems, and an instinct for where performance is being left on the table.
Experience working in cross-functional environments alongside ML researchers, infrastructure engineers, or applied scientists is a plus - translating requirements across disciplines without losing rigour.
Familiarity with large-scale feature normalisation, sequence modelling infrastructure, or foundation model serving is a genuine differentiator in this role.
Show more Show less