Artificial Intelligence Engineer

London | Office-first (4 days/week)

A well-funded, science-led AI company is hiring a Senior AI Engineer to join its core Science team as it scales a platform that builds synthetic populations capable of predicting real-world human behaviour.

The company operates at the intersection of behavioural data, statistical modelling and generative AI, building simulation systems that model collective human decision‑making. Its technology is already deployed in enterprise environments where insight quality, speed and robustness directly impact strategic outcomes.

This is not a generic LLM engineering role. It sits at the intersection of research, behavioural modelling and production AI systems — designing the cognitive architectures that power realistic synthetic agents, and working closely with engineering to scale them reliably.

As a Senior AI Engineer, you will sit within the Science team, working on the research, design and implementation of AI agent systems that simulate human behaviour at scale.

• Design and run structured experiments to evaluate and improve synthetic agent behaviour
• Develop and iterate on agent cognitive architectures (memory, reasoning, personality and context systems)
• Build sophisticated prompting and behavioural frameworks that enable realistic decision‑making
• Validate model outputs against real-world data using statistical and experimental methods
• Work closely with backend engineering to scale agent systems into robust production environments
• Balance research exploration with practical delivery in an evolving product environment

This role is ideal for someone who enjoys both thinking deeply about models and behaviour and building real systems that operate reliably in enterprise settings.

• Designing memory and context systems that allow agents to maintain coherent beliefs and behaviour over time
• Running behavioural experiments to test whether simulated populations align with observed real‑world outcomes
• Comparing prompting strategies, model variants or fine‑tuning approaches to improve behavioural realism
• Building multi‑agent systems where agents reason, debate, retrieve evidence and update beliefs
• Diagnosing behavioural drift and improving model robustness through structured experimentation

• Evidence of running experiments to validate model behaviour and performance
• Experience iterating quickly on model architectures based on empirical results
• Understanding of statistical validation and data quality assessment
• Exposure to research‑driven product development or academic AI research

• Hands‑on experience building applications on top of large language models
• Experience with advanced prompting, RAG systems and agentic workflows
• Exposure to multi‑agent systems, simulation frameworks or agent‑based modelling
• Familiarity with frameworks such as PyTorch, Hugging Face, Lang Chain or similar
• Strong Python skills with backend frameworks (FastAPI, Django, Flask or similar)
• Understanding of scalable architectures and distributed systems
• Comfort taking systems from prototype through to production deployment

• Strong ownership mentality — you see ideas through from concept to deployment
• Intellectually rigorous — you question assumptions and adjust direction based on evidence
• Comfortable working in ambiguity and fast‑evolving environments
• Collaborative and direct — able to debate ideas constructively
• Genuinely motivated by understanding how AI systems behave, not just wiring them together

• Competitive salary with meaningful equity
• Office‑first culture in central London (4 days/week, non‑negotiable)
• Direct collaboration with senior scientific and technical leadership
• The opportunity to work on AI systems that influence real‑world, high‑stakes decisions
• Long‑term growth within a company scaling both its applied research capability and product footprint