Senior Autonomy Engineer (all genders)

STARK is a new kind of defence technology company revolutionizing the way autonomous systems are deployed across multiple domains. We design, develop and manufacture high-performance unmanned systems that are software-defined, mass-scalable, and cost-effective. This provides our operators with a decisive edge in highly contested environments.

We're focused on delivering deployable, high-performance systems - not future promises. In a time of rising threats, STARK is bolstering the technological edge of NATO Allies and their Partners to deter aggression and defend Europe - today.

This is not a research role. We are looking for a Senior Autonomy Engineer to build the intelligence behind real UAV swarming - the algorithms that decide which vehicle attacks which target, how the swarm re-organises when conditions change mid-mission, and how 4+ loitering munitions coordinate autonomously to overwhelm adversary defences.

You will design and implement the swarm decision-making layer at the heart of Minerva Frontline, our Mission Management Software. This means building the optimisation and planning algorithms that sit above our existing mission planning engine, turning a set of operator-level objectives and battlefield conditions into coordinated, multi-vehicle strike missions—then dynamically re-planning them when the battlefield shifts.

This is Genesis-stage work. There is no off-the-shelf library for what we're building. You will work shoulder-to-shoulder with our Software Architect and our backend team to take swarm coordination from concept to live flight demonstration. If you want to solve hard multi-agent problems that have real consequences in the physical world, this is the role.

• The Swarm Brain: You will design and implement the task allocation engine that optimally assigns multiple vehicles to multiple targets - accounting for threat priority, approach geometry, timing constraints, and vehicle capabilities. Think auction-based allocation, Hungarian method, and constraint satisfaction—applied to coordinated strike.

• Dynamic Re-Planning: You will build the system's ability to adapt mid-mission. When ISR reveals a target has moved, when a vehicle is lost, or when new targets appear, your algorithms re-evaluate the situation and generate updated plans - pushed to vehicles in flight.

• The Optimisation Layer: You will build a new planning layer that sits above our existing static Mission Plan system. Your layer decomposes swarm-level objectives into individual vehicle assignments, then delegates path generation to the existing engine.

• Simulation-First Development: You will iterate rapidly in our SITL (Software-in-the-Loop) simulation environment, testing swarm behaviours with 4+ virtual vehicles before anything flies. You will define the metrics that tell us whether the swarm is making good decisions.

• An Autonomy Builder: You have 5+ years of experience building autonomous decision-making systems - whether in robotics, aerospace, defence, automotive, or multi-agent AI. You have taken planning or coordination algorithms from research to working code in production or field environments.

• Mathematically Fluent: You are comfortable with combinatorial optimisation, graph theory, constraint satisfaction, and multi-agent coordination. You can read a paper on CBBA (Consensus-Based Bundle Algorithm) and have a working prototype by end of week.

• A Strong C++ Engineer (non-negotiable): You write production-quality Modern C++. Python is fine for prototyping and simulation scripting, but the production planning engine is C++ and you must be comfortable owning that codebase. You understand that your algorithms must run in a real-time C2 system where latency matters and reliability is non-negotiable.

• Systems-Minded: You don't just write algorithms in isolation. You understand how your planning layer integrates with the rest of the system - state management, gRPC services, telemetry feeds, and operator UX. You care about the interfaces as much as the internals.

• High Agency: You don't wait for perfect specifications. You work with domain experts to understand the operational need, define the problem mathematically, and build and iterate.

• Field-Ready: You are willing to support live flight demonstrations and field trials, working alongside integration and hardware teams to prove that your algorithms perform under real-world conditions.