O

Oxford Dynamics

Generative AI and autonomous defense robotics

Senior Applied AI Engineer

Natural Language Processing & LLMsConfirmed live in the last 24 hours
Contract£700 - £900

About the Job

Requirements

  • Significant professional software engineering experience, typically 6+ years, including several years building LLM-based or agentic systems, with a demonstrable track record as a senior or lead engineer.
  • Strong software engineering fundamentals, including writing clean, tested, production-grade code and making sound systems-level design decisions rather than building only working prototypes.
  • Commercial experience building multi-agent or agentic AI systems in production and owning them as an engineer rather than integrating them as a user.
  • Strong Python skills and hands-on experience building on LLM frameworks such as LangGraph, LangChain, or Haystack, with the ability to work below the framework when needed.
  • Experience designing and deploying AI/ML systems into real production environments rather than only notebooks or demos.
  • A solid grasp of engineering around models, including inference, latency and performance, data pipelines, state and memory, and tool or function calling.
  • Hands-on experience building search and retrieval over very large multimodal datasets of 3TB+ of text, imagery, telemetry, and sensor data, including indexing, embedding, querying at that scale, and feeding results into LLM or agentic pipelines with useful latency and precision.
  • Fluency with Docker, Git, and cloud platforms; AWS is preferred.
  • Understanding of secure deployment patterns, including air-gapped, on-premises, or sovereign cloud.
  • A builder's mindset, with the ability to ship dependable systems and explain them clearly to technical and non-technical people.
  • Senior technical ownership, including owning delivery end-to-end, taking technical ownership of the workstream, setting the direction other engineers build to, and ensuring disciplined documentation and knowledge transfer.
  • Applicants must be eligible for UK security clearance and hold a UK passport with five years of continuous UK residency. Applicants who are not eligible cannot progress.

Responsibilities

  • Design and build multi-agent AI systems from the ground up, using frameworks such as LangGraph or Haystack where useful and building the orchestration, state, and tool-calling infrastructure underneath when needed.
  • Write clean, tested, maintainable code and make sound systems-level decisions, treating reliability, latency, and failure behaviour as first-class concerns.
  • Integrate large language and vision-language models into reasoning, search, summarisation, and task-execution pipelines, and build the retrieval, memory, and guardrails needed for dependability.
  • Deploy systems to cloud, on-premises, and, where required, fully offline environments, engineering for each environment's security and reliability constraints.
  • Build production pipelines from data ingestion through inference and own the whole path.
  • Build evaluation, observability, and guardrails that demonstrate system behaviour, including agent behaviour, model performance, and failure modes.
  • Act as a technical authority for agentic AI, take technical ownership of the workstream, set design patterns and direction for other engineers, and leave maintainable systems and documentation after the engagement ends.

Desired Qualifications

  • Experience with multimodal reasoning.
  • Experience with edge or offline AI deployments.
  • Familiarity with Kubernetes, including EKS or OpenShift, for managing deployed applications.
  • MLOps experience, including model evaluation, monitoring, and reproducibility.
  • Observability tooling for agentic systems, including model drift, agent behaviour, and performance monitoring.
  • Experience with agent orchestration patterns and inter-agent communication protocols such as A2A.
  • Familiarity with the Model Context Protocol (MCP) for tool and context integration in agentic systems.
  • Familiarity with secure-by-design development principles, including ISO 27001, NIST, and OWASP.
  • Experience in defence, national security, or similarly regulated environments.
  • Contributions to open-source AI/ML projects.
  • An active SC or DV clearance.

About the Company

Oxford Dynamics

Generative AI and autonomous defense robotics

Oxford Dynamics builds sovereign AI and robotic systems for space, security, and defence, developing core IP for use on the ground and in orbit. Its AVIS platform provides offline, on-premises generative AI to search, analyze, and summarize data; SR-1 fuses multi-source sensor data to speed up the OODA-style decision cycle; Strider is a semi-autonomous robot for hazardous environments that can operate beyond the operator’s line of sight. The company differentiates itself by delivering on-premises, integrated AI and autonomous capabilities tailored to defence needs, and by pursuing both terrestrial and space applications through government contracts and strategic partnerships (including BAE Systems). Its goal is to unlock underutilized data for defence and national security, reduce human risk in dangerous missions, and build secure, sovereign technology through long-term collaborations.

Company size
11-50
Company stage
Not listed
Total funding
Not listed
Headquarters
Harwell, United Kingdom
Founded
2019

Recently Posted Jobs

Explore more opportunities at Oxford Dynamics

See more jobs