AI engineer in risk analytics in London

Would you like to apply the leading edge of AI to some of the hardest problems in risk and insurance, inside a global company that works with 91% of the Fortune Global 500? We are looking for an exceptional engineer to join our Willis Neuron Technology team in New York or London as Lead AI Engineer: a senior, high-autonomy individual contributor role in technology delivery, for someone already working at the frontier of AI engineering.

Responsibilities:

• Work as a forward deployed engineer within our delivery team: get close to brokers, actuaries, modellers and clients, and build AI solutions that differentiate WTW in the market, wherever they add the most value across R&A and Data.
• Prove ideas fast, with working software in days, then industrialise what works: take it through our delivery pipeline (backlog, code review, CI/CD, QA and release) into secure, supportable production services, handed over cleanly to service teams.
• Design and build agentic analytics and AI-enabled products using large language models, RAG, MCP and modern agent frameworks, as reusable components and services on our shared platform rather than one-off solutions.
• Own the quality bar for AI in production, including evaluation, testing, guardrails, observability, security and cost management, so that AI answers over risk data can be trusted by clients and colleagues.
• Bring analytical rigour and domain depth: apply your quantitative background and your understanding of insurance, exposure, loss and catastrophe modelling to build solutions that stand up to scrutiny from actuaries and modellers.
• Deliver as part of a team: plan and prioritise with Delivery Leads and Product Leads, and shape the technical direction for AI with Product Architects, within our architecture, security and compliance standards.
• Direct AI coding assistants and agentic workflows (e.g. Claude Code, GitHub Copilot) alongside your own hands‑on expertise to multiply your output, and rigorously review human-written and AI-generated code against our standards.
• Lead by example, without line management: raise the standard through the quality and pace of your own work, mentor colleagues informally, and share what you learn so capability spreads across the team.
• Stay at the frontier: track new models, tools and techniques, test them hands‑on, and bring the credible ones into our roadmap and delivery approach.

Requirements:

• 10+ years of professional software engineering experience, in any modern stack, with a record of leading complex delivery as a hands‑on technical authority and mentoring others through your work.
• Recent hands‑on experience (2+ years) building and shipping LLM‑based and agentic systems to production as part of a delivery team, using RAG, prompt engineering and agent frameworks / orchestration (e.g. LangGraph, Semantic Kernel), including evaluation, guardrails and cost management.
• Expert, hands‑on use of AI coding assistants and agentic development tools (e.g. Claude Code, GitHub Copilot, Cursor) to deliver exceptional personal output, with the discipline to critically review, validate and secure what they produce.
• Strong engineering capability in Python and at least one of C# / .NET or TypeScript (Angular and/or React), with hands‑on commercial experience of cloud (Azure preferred), GitHub or Azure DevOps, and the SDLC, CI/CD and Agile delivery.
• A strong analytical and quantitative background (e.g. mathematics, statistics, engineering or computer science), with the ability to reason rigorously about data, models and uncertainty.
• Strong domain knowledge of insurance, reinsurance or risk analytics (ideally actuarial or catastrophe modelling), or a record of becoming expert quickly in a complex quantitative field.
• Proven ability to work directly with business users or clients and move a concept from conversation to working demo in days, then carry it into production, challenging established approaches constructively and communicating complex trade‑offs clearly to senior stakeholders, actuaries and engineers alike.
• Proven ability to work well both individually and as part of a team: self‑directed and highly productive on your own, and collaborative in how you deliver with others, through shared backlogs, code review and handover to service teams.
• Evidence of staying current at the frontier of AI: recent tools, models and techniques you have tested hands‑on, and what you concluded about them.

Desirable:

• Experience designing multi‑agent workflows and building reusable MCP servers or tools that expose enterprise data and calculations to AI agents.
• Experience running AI / LLM solutions at scale in a regulated environment, including evaluation frameworks, model governance and responsible‑AI practices.
• Experience with our platform stack (C# / .NET, Angular or React, Azure), microservice architectures and containerisation.
• Experience with geospatial, catastrophe or financial‑modelling data, or with building mathematical algorithms.
• Earlier applied machine learning, statistics or data science experience, alongside your recent work with LLMs.
• Visible contribution to the AI engineering community, such as open‑source work, writing or speaking.
• Degree‑level qualification in a quantitative discipline (Mathematics, Science, Engineering or Computer Science); postgraduate study valued.

Conditions:

• The role offers hybrid working with travel to office expected on a regular cadence for meetings and workshops. Some occasional travel may be required as part of the engagement with business and technology stakeholders.
• The base salary compensation range being offered for this role is $100,000– $350,000 USD per year. This role is also eligible for an annual short‑term incentive bonus.
• WTW provides a competitive benefit package which includes health and welfare benefits, leave benefits, and retirement benefits.