Principal AI Engineer

Government Recruitment Service; Greater London; Hybrid; GBP 54,000 - 66,000; Full time

Benefits offered by this job

Disability Confident

Job summary

Government Recruitment Service is seeking a Principal AI Engineer in London. You will lead the technical delivery of AI and data science products, design and deploy scalable AI solutions, and collaborate with government teams to ensure responsible adoption across transport systems.

You will own architecture decisions, support production readiness, and drive MLOps, testing, and governance across projects. Hybrid work is available.

Qualifications

• Leading end-to-end technical delivery of AI/data science products in production environments.
• Programming and software engineering for scalable solutions with proper engineering practices.
• Deploying and managing products on cloud platforms using CI/CD pipelines and IaC tools.
• Applying a range of AI technologies (LLMs, VLMs, RAG, Agents, ML) in production.
• Collaborating across multidisciplinary teams to translate priorities into technical solutions.
• Statistical evaluations of AI solutions using appropriate designs and datasets.
• Considering ethics and GDPR in AI solutions and human oversight design.
• Using AI architecture expertise to advise on design choices and reusable patterns.
• Maintaining multiple AI products in production with MLOps practices.
• Building agentic services using enterprise platforms like Copilot and Gemini Enterprise.
• Red-teaming AI products to uncover vulnerabilities and mitigations.
• Creating intuitive front-end interfaces for users.
• Training, fine-tuning and deploying models for specialist tasks.

Responsibilities

• Lead end-to-end technical delivery of AI/data science products in production environments.
• Design, build, deploy, monitor and maintain AI solutions across the lifecycle.
• Collaborate with DfT and external partners to implement AI responsibly.
• Deploy products across development, test and production with cloud IaC tooling.

Skills

End-to-end delivery

AI architecture

Cloud deployment

MLOps

Collaboration across teams

Statistical evaluation

Ethics & GDPR

CI/CD pipelines

Front-end interfaces

Red teaming

Model training & fine-tuning

Tools

Terraform

Terraform or IaC tools

CI/CD

GCP

AWS

Copilot

Gemini Enterprise

Job description

Principal AI Engineer

Company: Government Recruitment Service Salary: from £59,816 a year Hours: Full-time Location: London Working pattern: Hybrid - 2 days remote Job type: Permanent Posting date: 4 Nov 2026 Closing date: 15 Nov 2026

Summary

Do you want to develop impactful AI products that address real-world transport challenges?

Are you an innovator who is driven to implement AI within DfT to support our business and responsibly deploy it across the transport system to improve transport user experiences and boost growth?

We’re looking for Principal AI Engineers to lead the technical delivery of innovative, well-designed solutions that inform high-stakes decisions, transform complex processes and operations, and support critical policy objectives.

You will use your technical expertise to design, build, deploy, and maintain AI solutions – providing technical ownership through all phases of the product lifecycle. This involves effectively collaborating with colleagues within DfT, across government and with external partners.

Our Team

You will join the AI and Data Science Team, which owns the technical development of AI products and solutions in DfT’s AI Programme. Our AI Programme is split across five G6-led partner team(s): AI and Data Science own the technical development of AI and data science products (i.e., where this role sits); Human Centred AI design and implement social, behavioural and user research; Strategy, Governance and Responsible Adoption; and Transport sector AI Policy. We are also supported by the Problem Structuring and Analysis team. Our teams work together very closely.

As demand for our work has grown, we are now looking for new members to join us!

Essential requirements

• Leading the end-to-end technical delivery of impactful AI or data science products, including design, development, deployment, monitoring, assurance, incident response and continuous improvement in production environments (lead criterion).
• Programming and software engineering to design and develop high-quality, scalable and maintainable solutions, applying strong engineering practices including version control, testing and code review.
• Deploying and managing products across development, test, and production environments on cloud platforms (e.g., GCP/AWS), using CI/CD pipelines with automated testing and Terraform or similar infrastructure-as-code (IaC) tools to ship changes safely, reproducibly, and with rollback.
• Applying a range of AI technologies, such as LLMs, VLMs, RAG, Agents, MCP, computer vision, GraphML, recommendation systems or reinforcement learning.
• Collaborating across multidisciplinary teams, translating organisational priorities into technical solutions, communicating complex information effectively, and helping build organisational capability through setting standards, coaching and establishing reusable engineering practices.
• Statistical evaluations of AI and data science solutions using appropriate designs and datasets, applying lessons from research and practice.
• Considering ethics and legislation when developing AI and data science solutions e.g., GDPR, design of appropriate human oversight in AI tools.
• Using your credibility and expertise in AI architecture to advise and challenge colleagues on architecture design choices, ensuring proposals align with the organisation's technology landscape and strategy, establishing reusable patterns and working closely with the DfT Architecture team to support proposals through technical governance processes.
• Maintaining multiple AI products in production using MLOps practices, including monitoring, versioning, automated evaluation and CI/CD.
• Building and deploying agentic services using enterprise platforms such as Copilot and Gemini Enterprise, integrating backend services, APIs and organisational data to deliver secure, scalable solutions at pace.
• Red teaming AI products using off-the-shelf tools and custom-built tests to systematically uncover vulnerabilities and failure modes such as prompt injection, jailbreaks, data leakage, harmful outputs, and misaligned behaviour, and implementing effective mitigations.
• Building intuitive, front-end interfaces that users love.
• Training, fine-tuning and deploying models for specialist tasks.

Proud member of the Disability Confident employer scheme