AI Engineer

ICAP•New York, NY

11d•$183,600 - $222,000•Hybrid

About The Position

This is a newly created Phase 1 opportunity in our Global Broking division for an AI Engineer, joining the firm's new AI Solutions Group. Two engineers are being hired in this phase — one in London and one in New York — to establish the operating model for embedded AI delivery on the broking desks. The AI Solutions Group is a small, business-literate engineering team that sits alongside the brokers and embeds AI capability directly into the broking workflow — from the first client conversation through research, execution and booking. The team reports organisationally into Technology but spends its working day on the desks. Work is prioritised by Product Management against firm-wide value and reuse, and delivered on a Kanban model rather than a fixed quarterly cycle. As one of the two Phase 1 engineers, you will be embedded on a small number of pilot desks in your region. The objective of Phase 1 is to prove the operating model: that the team can ship AI solutions safely into production via the agreed governance route, and that brokers will adopt and use what is built. Every capability you build is placed into a shared library — designed to be queryable by an AI agent — so it can be reused across desks and regions as the team scales to its Phase 2 footprint. This role suits a strong engineer who is current on AI tooling and willing to sit on a desk with brokers, understand what they do all day, and translate that understanding into production-grade tools. You will work within the same Risk, Architecture, Technology and Operations standards as the rest of the firm, contributing to the safe route to production that the team consumes repeatably as it grows.

Requirements

Strong production engineering background — track record of architecting, building and operating scalable, robust services in production environments, not just prototypes.
Hands-on experience working with Large Language Models (LLMs), Generative AI and modern AI/ML frameworks; current on the AI tooling landscape and able to keep up as it evolves.
Expertise in Python in AI/ML frameworks and libraries.
Demonstrated ability to understand and engage with a complex business domain — capable of sitting with non-technical professionals, learning how they work, and translating that understanding into production tools.
Experience managing technical priorities on a Kanban or Agile backlog, resolving dependencies, and aligning delivery with both business value and centrally-set technical standards.
Evidence of presenting complex technical concepts to internal professionals and business stakeholders with varying technical backgrounds, tailoring style to the audience.
Demonstrated ability to deliver within a regulated environment — comfortable working through controls covering security, compliance, model risk, change management and audit, rather than treating governance as an obstacle.

Nice To Haves

Prior exposure to financial markets, broking, trading or other front-office environments is a strong plus.
AWS experience, including AI services such as Bedrock and the Serverless ecosystem, with the ability to work within established cloud frameworks.
Proficiency in front-end technologies, particularly React, with evidence of building intuitive, responsive user interfaces that integrate seamlessly with AI-driven back-end services.
Expertise in back-end development, including building and deploying scalable APIs and microservices to support AI solutions.
Knowledge of OAuth (Okta) for implementing secure, scalable authentication and authorisation in full-stack applications, ensuring data protection and privacy across AI solutions.
Experience contributing to or consuming a shared component / capability library across multiple teams or regions.
Prior experience embedding directly with business users — front-office desks, trading floors, advisory teams — rather than working purely from a central engineering function.

Responsibilities

Embed on assigned pilot desks in London or New York — sit with brokers, understand their workflow, and identify areas that are slow, repetitive or error-prone, as well as opportunities where AI changes what is possible.
Architect, design and build production-grade AI solutions — including Generative AI and agentic capabilities — that remove friction from the broking workflow and unlock new ways of working across idea generation, research, pricing, execution, booking and post-trade.
Deliver every solution through the firm's agreed safe route to production — covering security, compliance, model risk, change management, audit and operational ownership in production — with zero ad-hoc exceptions.
Contribute to and reuse from the shared AI capability library — check existing capabilities before building new ones, and place everything you build into the library so other desks and regions can adopt it without rebuilding.
Co-develop the team's shared platforms, frameworks and components with peers in the AI Solutions Group and the wider Technology organisation, customising for domain-specific desk requirements while preserving reuse.
Operate on a Kanban delivery model — work flows continuously, prioritised by Product Management against firm-wide value, with no individual desk independently commissioning its own tools.
Maintain strict data stewardship, model risk and compliance standards as AI agent and decision-support solutions scale across the broking lifecycle.
Provide ongoing support for deployed AI solutions, ensuring smooth operation and rapid resolution of issues affecting the desk.
Translate stakeholder requirements into technical solutions, ensuring alignment with both business objectives and the firm's enterprise technical standards.
Present project updates and key developments to broking leadership, Technology and governance stakeholders — demonstrating value, adoption and progression through the team's tiered scope (productivity, advisory, decision-making).