Head of AI Platforms & Deployment

Our client is a savings and investment app.

About our client

Our mission is to give everyone the means to get more out of life. We're guided by our belief that wealth isn't about the money, it's about the means to more - more freedom, opportunities, possibilities, and peace of mind. Our client 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.

Job Brief

Our client engineering serves more than 2M customers and runs a live service handling over 20M API requests a day. We have agreed a company-wide AI Platforms strategy, backed by a business case, and we are now building the team to deliver it. This is a newly created function and you will be its first hire.

The Head of AI Platforms & Deployment is a player-manager role reporting to the Engineering Director, owning both sides of that strategy: the AI Platforms capability pillars (control, monitoring and sandboxing; agentic workflow orchestration; knowledge capture) and the AI Deployment team of forward-deployed engineers who put AI to work on real business problems. You will be hands-on from day one - enabling business deployment of AI with the tools we have today - whilst also building and delivering a business-driven, multi-year AI platforms strategy and hiring and managing the team that executes it.

Why now:

• AI offers our client significant opportunities, and unlocking them safely means managing the risks that come with it. The controls and capabilities that make that possible (gateways, sandboxing, guardrails, monitoring, perimeter defence) need dedicated leadership to build.
• Demand for AI-enabled business systems is real and, without a safe path, goes underground. Someone must own the graduation ladder from "works for me" to owned, reviewed business system.
• We are hosting customer-facing AI models as part of our Aurora financial guidance service. Serving these safely, securely and at scale needs a dedicated engineering owner.
• We need to evolve our current Claude, Gemini and other frontier LLM deployments into a coherent, safe and effective strategic deployment platform.

You are not coming in to write a strategy deck. Pace and learning are key: we expect you to run two tracks in parallel from day one -

enable the do-ers

, helping the business use AI safely with what we have today rather than gating everything behind platform build, and

build the strategic platform

. Key decisions should be made within your first two to three months and material platform improvements live within six. You will hire further engineers before the end of the year, growing the team as the benefits compound.

What You'll Do

Own the AI platform strategy. Risk management, horizon scanning, supplier selection and vendor management, business case ownership, and benefits realisation through a coherent, business-driven delivery roadmap that is trusted by stakeholders across our client.

Build and lead the AI Deployment team. Hire, line-manage and set the engagement priorities and delivery standards for a team of forward-deployed AI engineers, working alongside embedded specialists while we hire. Grow into managing both the platform and deployment sub-teams as the function expands.

Own perimeter safety in an AI-native world. Set and own our defensive AI strategy, with execution carried out in collaboration with Tech Ops. Our Principal Cloud Architect, who owns overall cloud strategy, is a key partner.

Own AI platforms and costs. Harnesses, tooling, cost management, forecasting and optimisation (including usage-based pricing shifts), and the staff access model. This includes:

AI safety for internal use: sandboxing, policy, integrations, MCP governance, usage guardrails and monitoring expectations. The three platform pillars: control, monitoring and sandboxing (LLM and MCP gateway, anonymisation, local hardening, sandboxed cloud agent hosting); agentic workflow orchestration (a "software factory" in the cloud with hybrid self-serve and engineering-managed workflows); and knowledge capture and organisation. AI workflow tooling and the tiered graduation ladder from personal tool, to shared tool, to business system. Evolving existing deployments: tactical enhancement of what we run today so it aligns with the strategic direction.

Own model hosting and scaling. Hosting and scaling for AI and model workloads, including the customer-facing models behind our Aurora financial guidance service, with execution in collaboration with Tech Ops. Model safety and performance for customer-facing AI is shared with our Decisioning and Data Science teams: they own what happens inside the model, you own everything surrounding it.

Enable the do-ers. Give departments a working answer for using AI today - clear guardrails on what is allowed now and fast risk assessment rather than blanket restriction - and make sure demand arrives through the front door.

This role is explicitly not ML model development or data science, general cloud infrastructure ownership, or general engineering delivery - although our squads and engineering leads are customers of the platforms you build.

Who You Are

• A proven engineering leader who has gone deep on AI. You have managed engineers and managers, run programmes, owned budgets and supplier relationships, and in the last two to three years you have owned an AI platform or enablement capability in a real organisation. Not a spectator or strategist-only.
• A player-manager. You are comfortable getting hands-on in the early months and equally comfortable stepping back as the team you have hired takes it on.
• Strategic and accountable. You build and deliver a 12-18 month strategy aligned to company objectives, make well-informed and timely decisions, and take full ownership of your function's performance and commitments.
• Unusually cross-functional. Every department is your stakeholder and customer. You will be recognised as the AI platforms expert at our client, influencing exec and departmental leaders and resolving boundaries with Tech Ops and Decisioning through shared goals rather than turf.
• A team builder. You will build this team from scratch rather than inherit one, establishing a performance culture, developing your people, and running a team that itself models AI-enabled productivity.
• Commercially literate. You can express platform investment in ROI and EBITDA terms to a non-technical exec, and defend a risk position - on a connector approval, a token-versus-agent permission gap, a vendor - with clear reasoning.

Experience & Skill

Essential Proven engineering leadership: has managed engineers and managers, run programmes, owned budgets and supplier relationships. Has owned an AI platform or enablement capability in a real organisation in the last 2-3 years: deployed AI tooling company-wide, set governance, managed cost at scale. Has built or significantly scaled a team. Current, hands-on fluency with the modern AI stack: frontier model platforms, agentic tooling and harnesses, MCP and integration patterns, evals, prompt and context engineering. AI security and governance: sandboxing, gateways (LLM, MCP, AI), DLP, guardrail enforcement and data-boundary reasoning; able to form and defend risk positions. Cost engineering: usage-based pricing models, forecasting and optimisation levers. Enough software and cloud architecture depth to hold your own with Architects, Tech Ops and Decisioning, and to get hands-on with the platform in the early months. Our core stack is .NET on Azure and Python is the standard for AI, so we need polyglot pragmatism over allegiance to any single stack. Vendor and supplier evaluation and management in a fast-moving market. Evidence of business-case thinking: can express platform investment in ROI and EBITDA terms to a non-technical exec. Desirable Experience in financial services or another regulated industry, or a comparable risk-context environment. Hands-on experience selecting or implementing LLM/MCP gateways, sandboxed agent hosting or agentic workflow orchestration platforms. Experience serving customer-facing AI models in production and partnering with data science teams on model safety and performance.