C++ Trading Platform Developer

Dex; Greater London; Hybrid; GBP 120,000 - 180,000; Full time

Job summary

Dex's trusted partner is seeking an experienced C++ systems engineer to build and scale a high-frequency trading platform. You will own low-latency components, real-time market data ingestion, and order routing across numerous exchanges from a Linux environment in a hybrid London setup.

We value deep knowledge of modern C++, data structures, and low-level systems behaviour, with the ability to optimize performance and work directly with senior technical leadership.

Qualifications

• Strong hands-on mastery of modern C++ and STL internals.
• Proven experience building high-performance client-server apps on Linux.
• Deep understanding of CPU architectures, memory models, and networking stacks.
• Ability to reason about low-level systems behaviour.
• Based in or able to commute to London on a hybrid basis.

Responsibilities

• Build, scale, and optimize high-throughput C++ trading infra for real-time data capture and order execution across many exchanges.
• Profile bottlenecks in network I/O, memory allocations, and data structures to maximise performance.
• Architect robust client-server distributed systems and Linux data pipelines for growing trading volumes.
• Collaborate with CTO/leadership to shape architecture and roadmap, avoiding legacy constraints.

Skills

Modern C++

Linux networking

Data structures

Low-level systems

Analytical problem solving

London hybrid

Tools

Linux

Job description

This role is with one of Dex's trusted partner companies. We work closely with their teams to truly understand their culture, goals, and what they're looking for, so we can match you with the right opportunity and give you context about the role before you commit to a process.

The role

A top-tier algorithmic market maker providing critical liquidity across digital asset venues globally, operating at institutional scale with the execution speed of a lean tech startup. Backed by leading global tech investors and running on zero legacy architecture, their entire trading operation is powered by a high-frequency, proprietary platform built from the ground up to trade continuously across dozens of centralised and decentralised exchanges.

Reporting directly to the CTO, you will build and scale the core trading platform spanning real-time market data ingestion, low-latency order routing, connectivity, and telemetry. This is low-level systems engineering where hardware, network, and memory behaviour matter as much as distributed architecture. It is not an enterprise integration or maintenance seat: you will have direct platform ownership, zero layers of approval, and the scope to optimize systems where microsecond execution and absolute reliability directly impact the bottom line.

The work

• Build, scale, and optimize high-throughput, fail-safe C++ trading infrastructure handling real-time market data capture and order execution across dozens of exchanges worldwide.
• Profile and eliminate bottlenecks across network I/O, memory allocations, and data structures where performance trade-offs dictate execution viability.
• Architect robust client-server distributed systems and data pipelines on Linux to support exponentially growing trading volumes.
• Collaborate directly with executive technical leadership to shape the platform architecture and technical roadmap without legacy constraints or bureaucracy.

What You Bring

• Deep mastery of modern C++ and an exact mental model of how the standard library is implemented under the hood (both are tested rigorously in the interview).
• Proven track record building high-performance client-server network applications on Linux, with thorough fluency in core data structures and algorithmic trade-offs.
• Strong mechanical sympathy: genuine comfort reasoning about low-level systems behaviour, including CPU architectures, memory models, and networking stacks.
• Sharp analytical problem-solving ability, evaluated directly through technical assessments.
• Based in or able to commute to London on a hybrid basis.