Weapon Systems Algorithm Engineer

A weapon systems algorithm engineer writes and tests the mathematical logic that lets defence systems detect, track and engage targets — things like guidance laws for missiles, tracking filters for radar and infrared sensors, fire control calculations, and the decision logic that fuses data from several sensors into one picture. The work is mostly simulation and modelling on a computer: you build a mathematical model of the system and the world it operates in, run it thousands of times under different conditions, and work out whether the algorithm behaves correctly. You'd typically work for a defence prime contractor, a missile or sensor manufacturer, a specialist defence SME or a government research organisation, with the end customer usually the MOD or an allied armed force.

Starting salary: £30k–£38k
Live graduate jobs: Search all jobs
Main way in: Defence graduate schemes are the most common route
They usually ask for: A numerate degree is effectively mandatory

Approximate graduate salary

Roughly £30,000–£38,000 to start is a reasonable approximation for a graduate entering this field, with the higher end more typical of large primes, PhD entrants and roles in the South East. This varies considerably by employer, location and whether you join on a structured scheme or directly into a project team, and some employers add allowances linked to holding security clearance.

What you'd actually do

Building and running simulations in MATLAB and Simulink (industry-standard tools for modelling dynamic systems) — for example, modelling how a seeker tracks a manoeuvring target and seeing where the algorithm loses lock
Writing and debugging code, usually MATLAB, Python or C/C++, and sometimes converting a prototype algorithm into production code that will run on real hardware with tight memory and timing limits
Analysing trial data — recordings from flight tests, range trials or lab rigs — to compare what the real system did against what your model predicted, and investigating the discrepancies
Running large batches of simulation cases (Monte Carlo runs, meaning repeated randomised trials) to characterise performance across the full range of conditions a system might face, then producing plots and statistics from the output

Show 3 more

Writing technical reports and design justifications, because defence work is heavily documented and your analysis has to be traceable and reviewable years later
Reviewing requirements with systems engineers and arguing about what a requirement actually means in practice — a lot of early-stage work is turning a vague customer need into something mathematically testable
Sitting in design reviews, peer-reviewing colleagues' models and code, and occasionally supporting lab integration or trials, though hands-on trials time is a smaller part of the job than people expect

How graduates get in

Defence graduate schemes are the most common route. Large primes and missile/sensor manufacturers run structured two-year schemes with rotations across engineering disciplines, and you'd typically apply in the autumn of your final year for a September start.
Direct entry to a named graduate engineer or junior algorithm engineer post. This is common at smaller defence SMEs and consultancies, which often hire against a specific project need rather than a yearly intake.
Via a summer internship or industrial placement year with a defence employer. This is a strong route — placement students are often made offers before the main graduate round opens, and it also gets security clearance started early.
MSc or PhD entry, particularly for roles in guidance, navigation and control, radar signal processing or sensor data fusion. A PhD is not required but is genuinely common in the more mathematical corners of this field, and research-heavy teams sometimes recruit almost exclusively at this level.

Show 2 more

Sideways from a related first job — control systems, aerospace flight dynamics, signal processing in telecoms, or scientific software. Moving in after a couple of years elsewhere is realistic, though you'll need to go through clearance from scratch.
Government research and scientific civil service routes (such as defence science organisations) offer an alternative path, often with a more research-oriented flavour and different pay structures to industry.

What employers ask for

A numerate degree is effectively mandatory, but the exact subject matters less than you'd think. Aerospace, electronic, electrical or mechanical engineering, physics, maths, and sometimes computer science or cybernetics are all normal routes in. What matters is that you've done real mathematics — calculus, linear algebra, differential equations, probability — not just programming.
A 2:1 is the usual stated minimum at larger employers; some accept a 2:2 with a relevant master's or strong project work, and smaller firms are often more flexible. Some employers still ask for specific A-level grades including maths.
UK nationality and eligibility for UK security clearance. This is the single biggest practical gate on the role and it is non-negotiable for most posts. Many positions require Security Check (SC) clearance and some require Developed Vetting (DV), which involves detailed checks on your background, finances and residency history. Sole UK nationality is required for some roles, and employers are explicit about this in adverts.
Demonstrable programming ability, most often MATLAB, Python or C/C++. You don't need to be a software engineer, but you need to be comfortable writing code that does maths correctly.

Show 2 more

Evidence of applied technical work — a final-year project involving control, modelling, signal processing or simulation; a Formula Student or UAV team; or a relevant placement. Employers want to see you've applied theory to something messy and real.
Accreditation matters for later chartership: an IEng/CEng-accredited degree (MEng, or BEng plus a master's) makes the route to chartered status smoother, though it is not usually a hiring requirement.

Skills that matter

Control theory and dynamics

Guidance and autopilot algorithms are control problems, so understanding feedback, stability margins and state-space models is the core technical vocabulary of the job.

Probability and estimation

Sensors are noisy and targets are uncertain, so tracking and fusion algorithms rest on things like Kalman filters and statistical performance analysis.

MATLAB/Simulink and scripting

Almost all modelling, analysis and plotting happens in these tools, and fluency genuinely speeds up how many design iterations you can get through in a week.

Scepticism about your own results

A simulation will happily give you a confident, beautifully plotted wrong answer, and the engineers who do well are the ones who habitually sanity-check against physics and simple limiting cases.

Technical writing

Your analysis only counts if it survives a design review and can be audited later, so clear reports and well-justified assumptions are part of the deliverable, not an afterthought.

Working within constraints you didn't choose

Algorithms have to run on specific processors within fixed memory and timing budgets, so the elegant solution often loses to the one that fits.

Putting these on a CV? We counted which skills 12,914 live graduate adverts ask for, and what to do with the answer — part of our application advice.

Where it leads

Graduate or junior algorithm engineer, typically for the first two to three years: working on well-defined pieces of analysis under supervision, learning the toolchain, the domain and the documentation culture. Graduate schemes usually run around two years with rotations.
Algorithm or systems engineer owning a defined capability — a particular tracking filter, a guidance mode, a subsystem model — and being the person others come to with questions about it. Working towards chartered engineer status (CEng) through the IET, IMechE or RAeS is common at this stage, and many employers actively support it.
Senior engineer or technical lead, setting the technical approach for a workstream, mentoring juniors, and doing more of the customer-facing requirements negotiation. How long this takes varies a lot — programme length and the pace of hiring in defence mean some people get there quickly and others plateau for years on a long-running project.
From senior level the paths fork. Some go technical: principal engineer, technical authority, or chief engineer on a programme, which keeps you close to the maths. Others go managerial: project or engineering management, bids and proposals, or capability leadership. A few move into the MOD, defence research organisations or consultancy, where SC or DV clearance plus programme knowledge is highly portable.
Longer term, the skills travel reasonably well into civil aerospace, autonomous vehicles, space and robotics — anywhere that needs guidance, navigation and control or sensor fusion — though the clearance and the classified nature of your past work can make it hard to describe exactly what you did.

What people get wrong

“You'll spend your time around actual weapons, on ranges and in workshops.”

The overwhelming majority of the work is at a desk, in simulation. Many algorithm engineers go years without seeing the physical system they work on, and trials attendance is occasional rather than routine.

“It's a software engineering job with a military theme.”

It's closer to applied mathematics. The hard part is usually deciding what the algorithm should do and proving it behaves correctly across all conditions, not implementing it. Teams often have separate software engineers who take prototype algorithms into production code.

“The security clearance is a formality you sort out in week one.”

Clearance can take a long time and the timescale is outside your employer's control. New starters commonly spend their first months on unclassified work, background reading or training while it comes through, and some roles require sole UK nationality, which rules out otherwise strong candidates regardless of talent.

“Everyone has an aerospace engineering degree.”

Physicists and mathematicians are very well represented, sometimes more so than aerospace graduates, because estimation theory and signal processing are the harder things to pick up on the job. Domain knowledge about missiles or radar is taught in-house; mathematical maturity generally is not.

Where this varies

Practice differs a lot depending on who you work for. At a large prime contractor you'll likely specialise narrowly on one part of one long programme, with heavy process, formal design reviews and extensive documentation — stable, but slow-moving. At a defence SME or consultancy you may touch several projects a year and do your own coding, testing and customer liaison, with far less process but less structured training. Government research organisations lean more towards open-ended research and experimentation than delivery to a contract. Geography also clusters: much of this work sits in the South West, the South East, parts of the North West and central Scotland, around established defence sites, so relocating is often part of the deal. Finally, job titles are inconsistent — very similar work is advertised as guidance navigation and control engineer, sensor fusion engineer, signal processing engineer, modelling and simulation engineer or just systems engineer, so searching only for this exact title will miss most of the openings.

Search Weapon Systems Algorithm Engineer jobsFrom employers' own careers sites, updated daily Tailor your CV for this roleFree with an account, 3 a month

General guidance about the role across the UK market, not about any specific employer. Entry routes and requirements vary — always check the individual job advert.