AI Architect / Principal AI Engineer

We are seeking a hands-on Principal AI Engineer to act as the technical right hand for the Automation & AI Practice, accelerating the Colleague AI agenda. This pivotal role will transition the organisation from isolated AI experiments to a fully resourced, governed, and scalable AI delivery capability. You will seed in-house expertise to reduce vendor dependency, mitigate key person risk, and govern shadow AI usage across the business.

• Provide hands-on technical leadership across Colleague AI initiatives including Sierra PoC and Claude roll-out, owning architecture and design.
• Drive the build of agentic AI solutions for Tribeless colleagues.
• Consult on the broader Colleague AI backlog (complaint triage, helpdesk assistant, inbox routing, QA/QC), ensuring solutions are fit for purpose, secure, low risk, and cost-efficient amid consumption-based AI pricing.
• Define and embed AI engineering standards, reference architectures, and reusable patterns, collaborating with infrastructure teams to govern shadow AI usage.
• Lead AI literacy and enablement across Digital and the wider business.
• Mentor and uplift the Automation Practice team to diffuse AI capability and support transition to a permanent in-house AI Engineering Manager.
• Represent the Automation & AI Practice in cross-functional forums, translating AI strategy into delivery roadmaps and measurable outcomes., Sierra PoC - enhancing client NPS, helpdesk demand deflection, and colleague efficiencies.
• Claude Roll-out - boosting colleague productivity.
• AI Enablement - improving AI literacy and adoption across the business.
• AI Citizen Development Framework - accelerating solution time to market and reducing digital dependency.
• Copilot Continuous Improvement - driving ongoing colleague productivity gains.
  
  Strong expertise in Generative AI, LLMs, agentic architectures, and multi-agent systems with proven production delivery in regulated environments (FS preferred).
• Proven experience building agentic solutions using Microsoft Copilot Studio/Foundry, AWS Bedrock/AgentCore, and frameworks such as LangChain, LangGraph, CrewAI, Strands.
• Comfortable across Microsoft and AWS cloud ecosystems, including LLM gateway/Bedrock patterns, data residency, SSO/OIDC federation, and cost controls.
• Experience with Claude/Anthropic, OpenAI, Microsoft Copilot families, including MCP, connectors, skills, and trade-offs.
• Strong understanding of AI testing disciplines (accuracy, hallucination, bias, prompt evaluations) and AI governance/risk (DPIA, data classification, FCA considerations).
• Excellent stakeholder management skills, able to engage from board-level to engineer-level.
• Mentorship and capability-building mindset to strengthen the team and avoid key-person dependency.