Senior Applied AI Engineer; Private Sector Contractor

Role:
Senior Applied AI Engineer (Private Sector Contractor)

Rate: £700–£900 per day, dependent on experience

Location:
Hybrid: 2 days a week on site at Harwell, Oxfordshire; remainder remote (UK-based)

Engagement:
Contract, outside IR35: initial [6]-month term with strong likelihood of extension

Start:
Immediate Start

Oxford Dynamics is at an inflection point.

We build frontier AI for environments where the cost of a wrong decision is high: defence and national security, critical infrastructure, and commercial organisations making high-stakes decisions on complex data. We are growing fast, and the constraint on how fast is no longer ideas or opportunity. It is whether the systems we ship are good enough for the people who stake decisions on them.

That is where you come in. We need engineers who can build the systems this mission depends on, and build them to a standard that holds when the answer has to be right. You will be trusted to own real systems end-to-end, held to a single high standard, and judged on one thing above all: whether what you ship earns the confidence people place in it when the decision matters.

If you are the kind of engineer who builds rather than assembles, who wants your work measured by whether it holds up in production rather than whether it demos well, and who wants to see the direct line from your code to systems people depend on, we would love to hear from you.

Founded in 2019, Oxford Dynamics is a fast-growing UK frontier technology company developing both digital and physical AI systems built to operate in dynamic, mission-critical environments.

At the core of everything we build is AVIS™ (A Very Intelligent System), our orchestration platform: the intelligence layer that fuses multi-modal data, including text, imagery, telemetry and sensor feeds, so operators can interrogate complex information at speed and make better decisions under pressure. AVIS™ powers our digital decision products, such as ORION for multi-source decision-to-effect, and it powers our physical platforms.

We work where wrong decisions can be catastrophic. We partner with defence, security and private sector organisations internationally to bring frontier AI into the environments where it matters most, whether that is protecting nations, infrastructure and lives, or transforming how commercial organisations reason over their data and make high-stakes decisions.

As a Senior Applied AI Engineer at Oxford Dynamics, you will build and ship agentic AI systems in production: designing the multi-agent pipelines, the reasoning and retrieval, the model integration, and the infrastructure that carries them into production. These are systems people rely on to make consequential decisions, so what you build has to earn its confidence honestly and hold up under scrutiny.

This is a builder's role, not an integrator's, and it is unambiguously senior. We need an engineer who designs and implements systems rather than wiring together off-the-shelf parts, who takes technical ownership of the workstream and sets the direction other engineers build to. You will make systems-level design decisions, write production-grade code, and own the hard engineering that makes agentic AI reliable in demanding production environments, including secure, on-premises and offline deployments, from architecture through to deployment.

That ownership is real. You will own features end-to-end and be accountable for how they behave under pressure, not just whether they work on your machine. Reliability, latency and failure behaviour are your concern rather than an afterthought, because in the environments we deploy into, reliability and trust matter more than speed to market.

Frontier here means dependable, not experimental. You need the engineering depth to build systems that keep working under real-world conditions, whether that is heavy load, degraded connectivity or a fully offline environment, and the judgement to know the difference between a demo and a system someone can stake a decision on.

• Build the systems themselves. You design and build multi-agent AI systems from the ground…