INGENIEUR TEMPS PARTIEL (H/F) in London

Moneybox is an award-winning wealth management platform, helping over one and a half million people build wealth throughout their lives, whether they're saving and investing, buying their first home, or planning for retirement.

We are building a personalisation system that helps match customers with the right financial pathway at the right moment, across relevance, guidance, advice and risk monitoring. Aurora, our AI-powered financial guidance system, sits at the heart of this stack, but it is only one part of a much larger system. The harder problem is working with low engagement signals, a limited customer data footprint, strict regulatory boundaries and the need for every decision to be correct, auditable and defensible.

The system has multiple interacting layers, including ranking, orchestration, policy translation and belief state management. Building each layer well is achievable. Building the whole system in a way that is robust, scalable and easy to iterate on is the challenge we are hiring for.

This is a Staff-level individual contributor role, reporting directly to the Director of AI and Decision Intelligence. You will work alongside a Senior AI Researcher, Principal Data Scientist, Senior ML Engineer, Senior Data Scientist and two ML Engineers. There is no line management expectation. This role is about technical leadership through hands-on contribution, architectural judgement and the quality of your reasoning.

You will work with ML researchers, data scientists and ML engineers to:

• Own the overall system architecture, understanding how components interact, where dependencies create risk and where production realities challenge theoretical designs.
• Make day-to-day architectural decisions, defining what gets built, how it is structured and the interface contracts between components, while partnering with the Director of AI and Decision Intelligence on major strategic decisions.
• Identify architectural risks early, ensuring short-term decisions do not create long-term constraints or unnecessary technical debt.
• Guide technical decision-making across the team, ensuring implementation choices remain aligned with the long-term architecture and product vision, even when delivery pressures favour short-term solutions.
• Act as a technical sounding board for complex design and systems challenges, helping the team make pragmatic decisions in areas with significant uncertainty or trade-offs.

You naturally look for structural solutions rather than local fixes, focusing on the root cause rather than the symptom. You enjoy writing code and see it as a core part of the role. You care about building systems that are not only effective, but also understandable by the people who will operate and evolve them over time.

Experience operating at Staff Engineer, Principal Engineer or equivalent scope within teams building AI or ML-powered products and systems. A track record of owning end-to-end system architecture, from design through to production, for complex AI or ML systems operating under real-world technical, product and operational constraints. Strong software engineering fundamentals. You write clean, maintainable and reviewable code in Python or C#/.NET, and understand why engineering quality matters as systems scale.

Deep understanding of the trade-offs involved in AI system design, including latency versus accuracy, trainability versus interpretability, modularity versus coupling, and engineering pragmatism versus theoretical elegance. Sufficient ML knowledge to engage credibly in discussions around model behaviour, evaluation approaches and system design. You do not need to be an ML researcher, but you should be able to understand research outputs and make sound architectural decisions about how they are deployed and integrated into production systems.

Experience building systems in regulated or high-stakes environments where decisions must be auditable, explainable and defensible. An interest in AI safety and a thoughtful approach to the risks, limitations and unintended consequences of automated decision-making systems.