Applied AI Engineer - Multi-Agent Systems (Hybrid)

Oxford Dynamics; Greater London; Hybrid; GBP 90,000 - 130,000; Full time

Benefits offered by this job

Flexible working hours

Hybrid working model

Company pension (NEST) with employer 4

Private Healthcare

29 days holiday + public holidays

Job summary

OxfordDynamics in London is seeking Applied Artificial Intelligence Engineers to join multiple positions. You will build and ship agentic AI systems in production, designing multi-agent pipelines and integrating large language models and vision-language models into real workflows for defence and national security customers.

This delivery role owns features end-to-end from design through deployment in constrained environments where reliability and security matter as much as speed to market.

Qualifications

• Commercial experience building multi-agent AI systems in production.
• Experience deploying AI/ML systems to production environments.
• Familiarity with secure deployment patterns (air-gapped, on-premises, sovereign cloud).
• Strong Python skills and hands-on experience with LLM frameworks.

Responsibilities

• Architect, build, and optimise multi-agent AI systems and pipelines.
• Integrate LLMs and vision-language models into agent workflows for reasoning and execution.
• Deploy AI systems to cloud, on-premises, and air-gapped environments.
• Build production-ready pipelines from data ingestion through inference.
• Collaborate with engineers and product leads to translate requirements into working systems.
• Act as a technical authority for agentic AI and set design patterns for junior engineers.

Skills

Python

LLM frameworks

Production deployment

Observability

Collaboration

Tools

LangGraph

LangChain

Haystack

Docker

Git

AWS

Kubernetes

Job description

OxfordDynamics in London is seeking Applied Artificial Intelligence Engineers to join multiple positions. You will build and ship agentic AI systems in production, designing multi-agent pipelines and integrating large language models and vision-language models into real workflows for defence and national security customers.

This delivery role owns features end-to-end from design through deployment in constrained environments where reliability and security matter as much as speed to market.