Overview
The Enterprise AI Architect will spearhead the AI architecture workstream for a significant transformation program within a renowned UK technology business. Working on a fully remote basis, this role focuses on developing a robust, safe, and well-governed AI infrastructure, including designing and evaluating AI gateways and defining security controls. This position involves collaborating closely with engineers and guiding the AI architecture from initial draft to a buildable design, while mentoring an internal successor throughout the process.
Responsibilities
- Review and refine the draft architecture to create a detailed multi-layer design leading to a build specification.
- Lead the evaluation and selection of AI gateways, exploring options like LiteLLM, Kong, and Azure APIM.
- Design the request pipeline, environments, egress controls, and manage Entra ID identity including agent identity.
- Establish supply-chain controls and define strategies for model provider, residency, and isolation.
- Develop a permission-aware grounding approach to ensure data visibility aligns with user entitlements.
- Own the data loss prevention (DLP) and guardrail policy, defining security control specifications and threat models.
- Oversee governance of use-cases from conception to production, establishing registries, release gates, and approved tools.
- Deliver reference architecture patterns for priority initiatives, such as AI-assisted software delivery and customer service agents.
Requirements
- Proven experience in enterprise architecture and Zero Trust frameworks.
- In-depth knowledge of Entra ID and conditional access mechanisms.
- Experience with egress and network controls within production environments.
- Familiarity with LLM gateways, model routing, and supply-chain control methods.
- Desirable experience includes running an LLM gateway in a production setting.
- Understanding of RAG over governed data and MCP security protocols.
- Experience conducting enterprise security reviews and chairing design forums.