Platform/Product Engineer (up to £150k)

Dex; Greater London; On-site; GBP 80,000 - 110,000; Full time

Job summary

Dex is partnering with a defence-focused company to build a robust logistics middleware. You will own the backend infrastructure, including data processing and ML pipelines, optimized for constrained on-prem hardware and battlefield-like environments.

The role emphasizes traditional ML over deep learning, with high ownership and the ability to adapt to ambiguous, evolving requirements as contracts roll in.

Qualifications

• 2–5 years building production software in platform or infra roles.
• Experience with classical ML and signal pipelines on constrained hardware.
• T-shaped: able to balance business, technical detail and ambiguity.
• Background in defence or finance/hedge funds building signal pipelines.
• NATO citizenship is a hard requirement.

Responsibilities

• Own backend infra for a logistics middle-layer, interfacing with NATO C2 systems.
• Design data processing and signal pipelines for intelligence feeds.
• Optimise network performance under environmental and adversarial constraints.
• Develop and deploy classical ML models for tactical routing decisions.
• Work with constrained on-prem hardware (16-20 CPUs, 64GB RAM per server).

Skills

Production software

Platform engineering

Infrastructure engineering

ML pipelines

NATO citizenship

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

This company builds a critical logistics middle-layer for armed forces. Their software tracks inventory, flags supply gaps, and optimises tactical routes, collapsing an 84-hour ammunition-order cycle into the 16-hour window where battles are decided. They're backed by leading investors and solving problems where the competition is literally pen-and-paper.

You’ll own the backend infrastructure that makes this possible: data processing, network optimisations under battlefield constraints, and classical ML pipelines. This isn't chasing state-of-the-art deep learning; it's about robust, battle-proven traditional ML running on constrained, on-prem hardware. Expect high ownership and ambiguous requirements as the first defence contracts land.

The work

• Own backend infrastructure for a logistics middle-layer, integrating with NATO command-and-control systems.
• Design and implement data processing and signal pipelines for intelligence feeds (drone observations, C2 inputs).
• Optimise network performance and routing decisions under severe environmental and adversarial constraints.
• Develop and deploy classical ML models that distil intelligence into tactical supply-routing decisions.
• Work with constrained on-prem hardware (16-20 CPUs, 64GB RAM per server), focusing on efficiency and reliability.

What You Bring

• 2-5 years building production software, with a solid stint in platform or infrastructure engineering.
• Experience with classical/traditional ML (signal pipelines, linear models) or a genuine desire to lean into this style on constrained hardware, rather than deep-learning hype.
• T-shaped: comfortable navigating business context, technical detail, and ambiguity with multiple concurrent priorities.
• Background in defence, finance/hedge funds (building signal pipelines), or work requiring security clearance.
• NATO citizenship is a hard requirement for this role.