C++ High-Frequency Trading Engineer - Ultra-Low Latency

Durlston Partners; New York (NY); On-site; USD 420,000 - 700,000; Full time

Job summary

Durlston Partners in New York is seeking a C++ Software Engineer specializing in high-frequency trading to build and optimise ultra-low latency infrastructure for live markets.

You will work across the full trading system stack, collaborating with traders and researchers to push performance to sub-microsecond levels while ensuring reliability and scalability in a fast-paced environment.

Qualifications

• Strong modern C++ (C++17 or later) with commercial experience in low latency or performance-critical systems.
• Deep understanding of multithreading, memory management, and systems-level performance tuning.
• Solid Linux development background and familiarity with networking protocols.
• Prior HFT or proprietary trading experience is a strong advantage.
• Bachelor's degree in Computer Science or similar.

Responsibilities

• Work across the full trading system stack, building and optimising the infrastructure for live trading at scale.
• Design and implement high-performance C++ systems targeting sub-microsecond and nanosecond-level latency.
• Optimise production code for speed, reliability, and throughput under live market conditions.
• Collaborate directly with traders, quant researchers, and fellow engineers on technically demanding problems.
• Work across the full software lifecycle in a fast-paced, results-driven environment.

Skills

C++ (C++17+)

Low latency

Multithreading

Linux development

Networking familiarity

Education

Bachelor's degree in Computer Science

Job description

Durlston Partners in New York is seeking a C++ Software Engineer specializing in high-frequency trading to build and optimise ultra-low latency infrastructure for live markets.

You will work across the full trading system stack, collaborating with traders and researchers to push performance to sub-microsecond levels while ensuring reliability and scalability in a fast-paced environment.