Forward Deployed Engineer (Applied AI)

London based, typically two days a week in the office or with customers, with flexibility around how that works. Some travel to the UAE.

This role sits with a global IT solutions provider standing up something genuinely different: an autonomous rapid prototyping pod for banking, insurance and fintech clients. A small, elite team that wins its own work, takes an ambiguous client problem, and turns it into a working AI prototype in four to five weeks, then hardens it into production. AI is already central to this business; this pod is about delivering it at a pace the industry hasn't seen from an enterprise provider.

You will embed directly with clients, working alongside them day to day to understand their workflows, assess their environments and data, spot where AI genuinely adds value, and turn ambiguous requirements into working prototypes. You own the technical relationship from discovery through the first delivery sprint, then preserve that context and hand over cleanly to the delivery pod who scale it into production. This is applied AI at the sharp end. You are the person who can sit with a bank, explain in plain terms what AI can and cannot do, and then go and build something real that proves it. Around half the job is communication, the other half is building.

Why This Role Stands Out

You are shaping what gets built, not implementing someone else's spec. By the time you're in the room the commercial conversation is done, so your job is the interesting bit: validating what can actually be built and proving it with a working prototype. The pod operates like a startup inside an established business. It is fully autonomous and deliberately protected from the usual drag: no L2 tickets, no side of desk duties, no meeting culture. You build, demo, iterate. The work is genuine applied AI for regulated financial services clients where output has to stand up to real scrutiny, with a modern AI stack and GPU backed inference behind it. There is a growing UAE dimension to the business too. Nothing is expected, but if working in or relocating to the UAE would ever appeal, they will back you to do it.

Key Responsibilities

• Embed directly with financial services clients, understanding their workflows, pain points and where AI can add real value.
• Assess client environments, infrastructure and data availability to determine what is buildable.
• Turn ambiguous requirements into working AI prototypes inside the client environment.
• Own the technical relationship from discovery through the first delivery sprint, preserving context into a clean handover to the delivery pod.
• Design and build LLM and agentic solutions using Python, LangChain and LangGraph, RAG pipelines and vector stores.
• Build tools using agent workflows with clean API contracts, deterministic responses and audit logging.
• Work with NVIDIA inference tooling such as NIM, NeMo and Triton for GPU backed AI workloads where needed.
• Run demos, gather feedback, troubleshoot live and iterate quickly with users and stakeholders.
• Translate technical AI concepts into clear business language and shape practical opportunities clients will actually fund.
• Harden prototypes enough to prove they are production viable, containerising with Docker and Kubernetes.

Ideal Experience

• Strong applied AI or machine learning engineering background, building LLM based solutions in Python.
• Hands on experience with LangChain and LangGraph, RAG pipelines and vector stores.
• A track record of turning ambiguous requirements into working prototypes in real client environments.
• Genuinely client facing, comfortable embedding with clients, translating AI into business terms and shaping opportunities.
• Solid software engineering fundamentals across REST APIs, Docker and Kubernetes.
• Financial services, banking, insurance or fintech experience, or other regulated environments.
• Experience with the NVIDIA inference stack including NIM, NeMo and Triton, and GPU backed workloads.
• Agentic patterns and tool or function calling.
• Background from a technical consultancy, a forward deployed, solutions or customer engineering role, or a data and technology firm serving financial services.
• Comfortable with startup pace and ownership rather than heavy process.