AI Architect
HybridLondon, England, United Kingdom
Job Summary
Design and own end-to-end AI client-specific solution architectures, including Generative AI, Agents, and traditional ML. Define enterprise AI architecture patterns, reference architectures, and reusable building blocks within the Microsoft ecosystem. Lead AI governance, security, and Responsible AI by design to ensure solutions meet regulatory and ethical standards. Provide architectural oversight across delivery teams, ensuring quality and scalability. Support pre-sales and discovery by shaping solution approaches, technical estimates, and delivery roadmaps. Advise stakeholders on AI feasibility, trade-offs, risks, and long-term sustainability. Build hands-on prototypes, reference implementations, and reusable accelerators to set credible technical direction. Act as a trusted advisor defining guardrails and ensuring AI delivers real business value across enterprise and public-sector programmes.
Required Qualifications
- Significant overall experience delivering AI and Data solutions in complex enterprise environments
- Minimum of 5 years hands-on AI delivery experience
- Proven experience delivering production-grade AI solutions end-to-end, from discovery through to live operations and ongoing model management
- Hands-on expertise across Microsoft cloud and AI ecosystem, including Microsoft Foundry and data platform
- Strong experience with integrating low-code Power Platform solutions into enterprise-scale solutions, including hands-on experience with Copilot Studio
- Solid understanding of Microsoft security, identity and compliance (Entra ID, RBAC, network isolation, data protection)
- Strong experience designing LLM-based solutions, including Retrieval-Augmented Generation (RAG), prompt engineering strategies and agent-based architectures, leveraging frameworks such as Semantic Kernel, LangChain, and Microsoft Agent Framework
- Deep practical experience with classical Machine Learning, including supervised and unsupervised learning, feature engineering, model selection, training pipelines, evaluation frameworks and production lifecycle management, equivalent in depth to GenAI expertise
- Proven ability to assess client use cases and make well-reasoned architectural decisions between classical ML and Generative AI approaches; and articulate clearly why
- Strong experience in cloud-native architecture, including API-based and event-driven designs, enterprise integration patterns and architectures focused on scalability, resilience and reliability
- Experience with DevOps, MLOps and LLMOps practices, including CI/CD pipelines, automated testing, environment strategies and operating LLM-based solutions in production, with a strong focus on model and prompt lifecycle management, observability and evaluation, reliability and cost management
- Strong understanding of Generative AI governance, operational controls and cost evaluation, including monitoring usage, performance and designing solutions that are supportable, auditable and sustainable at scale
- Strong experience embedding Responsible AI and governance into Generative AI solutions, covering safety, risk management, transparency, explainability, human-in-the-loop design and regulatory alignment
- Must hold at least two relevant Microsoft certifications (AI-901, AI-103, DP-800 AZ-305., AI-300), or be actively working towards them
- Candidates must hold, or be eligible for, UK Security Clearance
Desired Qualifications
- Desirable experience in integrating enterprise data platforms with AI solutions, including familiarity with Lakehouse architectures and data engineering principles
- Proven experience in a consulting role, with the ability to translate complex technical concepts into clear business outcomes, effectively engaging non-technical stakeholders and senior leaders
- Good experience of managing and mentoring junior AI engineers or consultants
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.