Forward Deployed Engineer III - Justice AI Unit - Sheffield

Government Recruitment Service; Sheffield; On-site; GBP 58,000 - 98,000; Full time

Job summary

The Ministry of Justice's Justice AI Unit seeks an experienced AI-enabled product lead to develop high-impact AI products used across the ministry. You will own end-to-end delivery—from identifying the problem to deployment and scaling adoption—and work directly with frontline staff in courts, prisons and probation to co-create solutions.

Based in London with regular travel, you will thrive in a fast-moving, mission-driven environment, shaping AI strategy while ensuring governance, security and

Qualifications

• Experience building and deploying AI-native software in production environments.
• Experience owning all parts of the software and AI stack, including frontend, backend, database, AI and ML.
• Experience building solutions like predictive models, LLM workflows, scheduling algorithms and data pipelines.
• Ability to work and thrive in complex organisational contexts.
• Up to date on the latest developments in AI.
• Entrepreneurial or startup experience.

Responsibilities

• Develop high-impact AI products, models and agents that will be widely used across the Ministry of Justice.
• Work directly in the field, such as in courts, prisons and probation, to understand user needs and operational challenges, and co-build with users.
• Own end-to-end delivery from identifying the problem, building the product, deploying it and scaling adoption.
• Work directly with frontline staff to understand their needs and co-develop solutions.
• Build and deploy AI products with real-world impact across courts, prisons and probation services.

Skills

AI-native software

Owner of AI stack

Predictive models & LLM workflows

Complex organisations

AI developments awareness

Entrepreneurial/startup experience

Tools

NextJS

FastAPI

Postgres

Langfuse

Terraform

Azure

Job description

Salary: £58,000 - 98,000 per year

Requirements:

• Experience building and deploying AI-native software in production environments
• Experience owning all parts of the software and AI stack, including frontend, backend, database, AI and ML
• Experience building solutions like predictive models, LLM workflows, scheduling algorithms and data pipelines
• Ability to work and thrive in complex organisational contexts
• Up to date on the latest developments in AI
• Experience with our stack, including NextJS, FastAPI and Postgres
• Experience with LLM observability and evaluation tools like Langfuse
• Cloud and infrastructure experience, such as Terraform and Azure
• Entrepreneurial or startup experience

Responsibilities:

• Develop high-impact AI products, models and agents that will be widely used across the Ministry of Justice
• Work directly in the field, such as in courts, prisons and probation, to understand user needs and operational challenges, and co-build with users
• Own end-to-end delivery from identifying the problem, building the product, deploying it and scaling adoption
• Work directly with frontline staff to understand their needs and co-develop solutions
• Build and deploy AI products with real-world impact across courts, prisons and probation services

Technologies:

• AI
• Azure
• Backend
• Cloud
• FastAPI
• Frontend
• LLM
• Langfuse
• NextJS
• Terraform
• Support
• Security
• Web

More:

We are the Ministry of Justices Justice AI Unit, a small, fast-moving and innovative team building and shipping products quickly across the Ministry of Justice. Our work focuses on using AI to enhance, not replace, human judgement so we can deliver better outcomes, faster services and greater fairness for all. This nationally based role has London as the home base, with candidates expected to relocate or travel regularly to London, at least once per week, in line with business needs. We offer flexible working options, learning and development, a culture that supports inclusion and diversity, and a Civil Service pension with an employer contribution of 28.97%, alongside annual leave, public holidays and season ticket advance.