Agentic AI Software Engineer

Roke; Woking, Romsey; In loco; GBP 70.000 - 100.000; Tempo pieno

Descrizione del lavoro

ROKE in the United Kingdom is seeking an experienced software engineer with a strong AI interest to design and build secure, agentic AI solutions for national security contexts. You’ll collaborate across AI agents, software engineering, cloud, cyber and data challenges to deliver dependable systems that make a real difference.

Based from Woking or Romsey, you’ll help teams adopt effective agentic harnesses, tools and practices while upholding security, traceability and human accountability.

Competenze

• Professional Python development experience and strong software engineering fundamentals.
• Experience as an AI power user applying coding assistants or related tools to improve software delivery.
• Hands-on experience designing or building AI agents, tool-using applications or multi-step automations.
• Strong practices: Git, automated testing, CI/CD, observability and secure development.

Mansioni

• Redesign end-to-end workflows around AI agents, balancing autonomy and human oversight.
• Design, build and deploy secure AI agents and multi-agent systems for mission-critical software.
• Use agentic harnesses and AI-assisted development tools to accelerate engineering with quality and traceability.
• Evaluate off-the-shelf vs bespoke capabilities from Elasticsearch and AWS.
• Create reusable patterns, guardrails, evaluation approaches and delivery practices.

Conoscenze

Python development

AI power user

AI agents design

Git

CI/CD

Observability

Agent autonomy vs guardrails

DV clearance

MLOps

Strumenti

Elasticsearch

AWS

LangChain

LangGraph

Descrizione del lavoro

You’ll help design and build ethical, secure agentic AI solutions for national security teams. You’ll bring a strong software engineering background, use AI as a power user, and turn emerging agent capabilities into dependable systems that make a real difference.

Based from Woking or Romsey, you’ll work across agentic AI, software engineering, cloud, cyber and data challenges. You’ll act as a force multiplier for AI-assisted development, helping teams adopt effective agentic harnesses, tools and working practices.

What you’ll be doing

Redesign end-to-end workflows around AI agents, identifying where autonomy, human oversight and conventional software provide the best outcomes.

Design, build and deploy secure AI agents and multi-agent systems that integrate with mission-critical software, data and services.

Use agentic harnesses and AI-assisted development tools to accelerate engineering while maintaining quality, security, traceability and human accountability.

Evaluate when to use off-the-shelf agent frameworks and managed services, including capabilities supplied by Elasticsearch and AWS, and when a bespoke approach is justified.

Create reusable patterns, guardrails, evaluation approaches and delivery practices that help multidisciplinary teams adopt agentic AI safely and effectively.

What you’ll bring

We know great candidates may not tick every box. We’re looking for an experienced software engineer who is enthusiastic about AI, understands how agents change software delivery, and can help others use these capabilities effectively. The must-haves are below; the rest can be developed with support from the team.

Must-haves

• A strong software engineering background, with professional Python development experience and sound knowledge of architecture, APIs, testing and maintainable code.
• Experience as an AI power user, applying coding assistants, agentic harnesses or related tools to improve software delivery.
• Hands-on experience designing or building AI agents, tool-using applications, retrieval workflows or multi-step automations.
• Good engineering practices, including Git, automated testing, CI/CD, observability and secure development.
• The judgement to balance agent autonomy with guardrails, evaluation, human oversight, security and responsible AI principles.
• Eligibility and willingness to achieve DV clearance.
• MLOps – Integrating machine learning workflows with DevOps and data engineering practices.

Nice to have

• Knowledge of agent-building frameworks and managed capabilities such as those supplied by Elasticsearch and AWS, together with the ability to assess off-the-shelf options against bespoke solutions.
• Experience with LLMs, prompt and context engineering, RAG, tool use, memory, orchestration and agent evaluation using frameworks such as LangChain or LangGraph.
• Practical experience deploying containerised, observable and secure AI solutions on AWS.
• Relevant certifications e.g. AWS Certified Machine Learning Engineer – Associate, AWS Certified Generative AI Developer – Professional.

Why ROKE?

At ROKE, you’ll work with smart, welcoming teams on technology that helps protect lives and national security. You’ll have the opportunity to shape how agentic AI transforms engineering and operational workflows, supported by mentoring, coaching and the ROKE Academy.

Competitive salary, plus benefits and bonus (dependent on skills and experience).

You’ll also have 30 days’ holiday plus bank holidays, private medical insurance, and health, wellness and family support - plus other benefits.

Relocation packages are available for suitable candidates.

Where you’ll work

Gloucester - You’ll find our Gloucester site in a business park two minutes from junction 11A of the M5; The site allows easy access to our local customer base. Set on the outskirts of the Cotswolds, you are never far from a picturesque view or lunch time walk.

Clearance

This role requires you to be eligible and willing to achieve DV clearance. Eligibility criteria are available on the gov.uk website.