Forward-Deployed AI Engineer

Investment Bank

You will own the journey from discovery through implementation: understand how work actually happens, identify where AI can materially improve it, rapidly put solutions in front of users, and partner with senior Global Markets quants and technology teams on governance, productionisation and scale.

This is a role for someone who likes interpreting and doing, not consulting and enabling from the sidelines.

What you will do

• Embed with Distribution desks. Understand how work actually happens, map workflows end to end, identify friction and waste, and prioritise opportunities against the seven pillars of the Global Markets Sales strategy.
• Build quickly. Use the bank's existing AI and low-code capabilities to automate workflows and get useful solutions in front of users in days rather than months. The objective is not to build bespoke software where an existing tool can solve the problem.
• Redesign before you automate. Challenge unnecessary steps, hand-offs and duplication rather than simply using AI to automate an inefficient process.
• Own the dependencies. Identify the APIs, data and entitlements required for your solutions, specify what is needed clearly, and drive those requirements through to delivery.
• Design for compliant AI. Understand the architectural and governance considerations around AI systems and ensure solutions operate within the bank's AI, data, risk and control frameworks.
• Unlock the right data. Recommend how each desk should organise and access its information to enable AI use cases while adhering to the bank's need-to-know framework.
• Know what not to build. Solve problems directly where the available tools allow it. Where a requirement genuinely needs strategic engineering, route it to Quant or Technology with a well-defined problem statement, requirements and expected business outcome.
• Drive adoption, not just delivery. Train users, observe how solutions perform in real workflows, iterate quickly, spread successful patterns across desks, and make sure what you build is still being used a month later.

• Scale what works. Spot recurring problems across products, regions and desks and turn successful local solutions into reusable capabilities.