Senior AI Evaluation Engineer; Data Scientist

Applications for this position close on 3rd November 2026 - 23.45pm UK Time

Incubator for AI, i.AI

Help shape how people interact with some of the most ambitious AI products in government

The Incubator for AI (i.AI) is a team of ~80 builders and operators using frontier AI to improve public services across healthcare, education, planning and parliament. We combine startup pace and autonomy with the opportunity to tackle national-scale challenges.

We operate under three core principles:

Talent :
We bring together the UK's best AI talent across a range of functions. You will work alongside exceptional researchers and top government leaders, staying at the cutting edge of technology.

Innovation :
We set precedents for what is possible in government. We combine the pace of a start-up with the influence of being at the digital centre of government. You will test new ideas, expand what is possible, and leverage unique government data to create novel solutions.

Impact :
We are dedicated to using AI as a tool for public good - this can mean improving outcomes in schools, boosting house building or providing more personalised support for those in need. With the backing of the Prime Minister, you will turn technical breakthroughs into real-world applications that affect millions of citizens.

i.AI's Impact and Evaluation team build their work into each stage of product development, from scoping, to incubation, and onwards to scaling. Team members embed into product development teams, designing and delivering robust evaluations at each stage of the product journey to inform their design and, ultimately, understand their impact on the quality and efficiency of public service delivery.

The team is now recruiting for two Senior AI Evaluation Engineers, to lead the evaluation of AI products. These roles will focus on model-level evaluation, including assessment of accuracy, bias and safety.

• Lead the design and development of evaluation pipelines and experiments across AI products
• Architect product-wide evaluation suites that integrate with engineering and CI/CD workflows
• Design and implement complex and exploratory evaluation methods to assess accuracy, bias, safety and overall model performance
• Work closely with product, engineering and domain experts to shape evaluation approaches throughout the product lifecycle
• Design and run evaluation experiments that directly inform product development and technical decisions
• Analyse complex evaluation results and turn findings into clear recommendations for technical, product and senior stakeholders
• Lead the communication of evaluation findings through reports, presentations and external publications
• Support and review evaluation work from other engineers, helping to raise standards and develop evaluation capability across the team

You may be a strong fit if you bring some of the following:

• Broad experience of evaluating generative and predictive AI systems
• Extensive knowledge of experimental design and statistical hypothesis testing
• Highly proficient in Python
• Demonstrable experience of software development tools and practices (e.g. git, CI/CD, unit testing)
• Ability to communicate technical work to a broad audience, using different media/formats e.g. technical reports, presentations, and web-based tools.
• Experience implementing LLM-as-a-judge and other generative AI evaluation techniques using AI evaluation tools (e.g. pydantic evals, deepeval, inspect)
• Experience using pipeline orchestration tools e.g. DVC, snakemake
• Knowledge of standard quantitative research designs, methods and analysis (particularly those set out in the Magenta Book)
• Experience of mentoring junior colleagues

• Career-defining…