Data Science/Machine Learning Engineers - ML AI

Data Science/Machine Learning Engineers - AI Assessment Models

Initial 6 months | £dependant on experience per day (Inside IR35) | Likely extensions | Hybrid - London/Remote

Overview
We are supporting a large exam and qualification authority, in recruiting two experienced engineers to help build the next generation of AI-powered assessment tools. This is a pioneering project with the long-term aim of an AI model that can be used to support the quality assurance cycle.

You'll be working on large language model (LLM) solutions to:

• Automatically assess responses against marking criteria and mark schemes
• Moderate and validate outcomes
• Support the transition to AI-powered assessment tools and processes

With limited in-house data science capability, we are seeking people who can operate end-to-end - from data engineering and preparation through to machine learning model design, training, evaluation, and deployment.

Key Responsibilities

• Work closely with programme management to shape the AI assessment model strategy.
• Design, build, and train large language models (LLMs) for quality assurance of assessment materials.
• Develop robust data pipelines for ingesting, cleaning, and transforming large volumes of responses and marking data into model-ready formats.
• Implement evaluation frameworks to compare AI-generated outputs against human outputs for accuracy and fairness.
• Research and implement best practices for explainable AI (XAI) to meet transparency requirements in education assessment.
• Collaborate with non-technical stakeholders, translating complex ML/AI concepts into clear, actionable project updates.
• Test, validate, and iterate model performance prior to large-scale rollout.

Skills & Experience

Essential

• Strong background in data science or machine learning engineering, with end-to-end project delivery experience (Python, Tensorflow, PyTorch or similar)
• Hands-on expertise in large language models (LLMs), ideally including fine-tuning, and natural language processing (NLP).
• Proven ability to build ETL/data engineering pipelines (Python, PySpark, SQL, or similar).
• Strong grounding in data preparation, cleansing, and quality assurance for AI/ML projects.
• Experience applying ML to classification, scoring, or moderation problems.
• Ability to work in environments without established data science infrastructure - building from scratch where needed.
• Excellent communication skills, able to engage with technical and non-technical audiences.

Desirable

• Background in education technology or assessment models.
• Familiarity with AI ethics, fairness in ML, and bias detection.
• Experience deploying models to cloud environments (AWS, Azure, GCP).
• Understanding of explainable AI (XAI) in regulated sectors.

Contract Details

• Day Rate: £500 per day
• IR35 Status: Inside (pending client confirmation)
• Location: Hybrid - likely 1-2 days London office but negotiable, remainder remote
• Initial Duration: 6 months (highly likely to extend into 2027-2028)

Why This Role?

This is a rare opportunity to shape the future of assessment in the UK. You'll be at the forefront of introducing AI into one of the country's most important educational processes, with the freedom to design and build solutions from the ground up, and see them rolled out at national scale.