Lead AI Solutions Architect

Role

Lead AI Solutions Architect embedded in a business domain to act as the primary technical authority for AI. The role bridges high-level AI strategy and production delivery, designing scalable, secure, and resilient ML, Generative AI and agentic AI solutions while refining internal AI platforms and engineering standards.

Key Responsibilities

• Serve as the primary AI solutions architect and trusted advisor from ideation through production.
• Ensure enterprise data, ML and AI architecture principles are applied pragmatically to deliver business value.
• Partner with business stakeholders, product managers, and engineering teams to understand use cases and constraints.
• Design and document end-to-end AI system architectures, including reusable AI services, AI lifecycle management frameworks (MLOps, LLMOps, AgentOps), feature stores, RAG pipelines, and inference architectures.
• Review and guide solution designs for security, scalability, resilience, and cost-effectiveness.
• Work with domain leadership to shape data and AI strategy and roadmaps; represent the domain in central Enterprise Architecture and AI communities.
• Provide structured feedback to central teams to evolve enterprise blueprints, platforms, and governance.
• Represent the business in architectural review forums and ensure compliance with security, AI governance, and risk standards.
• Champion architectural best practices, data literacy, and responsible AI design while balancing governance with delivery enablement.
• Identify opportunities for data, ML and AI to deliver business value and advise on prioritisation and sequencing.

Requirements

• Deep knowledge of modern AI architectures on AWS, including Generative AI, RAG and large-scale asynchronous inference workloads.
• Expert-level experience designing and operating AWS and Databricks, with mastery of model serving, vector search, and integration of foundation models.
• Strong understanding of AI security architectures (private model endpoints, PII masking/redaction, IAM least-privilege, secure data egress/ingress).
• Experience with Infrastructure-as-Code (Terraform) and containerisation (Docker, Kubernetes, Helm), and designing for GPU-accelerated compute, distributed training and high-availability inference.
• Hands-on application of MLOps and LLMOps principles: automated model evaluation, drift detection, CI/CD for model code/weights, and operational resilience of real-time inference APIs.
• Experience with AI FinOps and cost-performance trade-offs between proprietary and open-source models (e.g., Llama, Mistral).
• Proven ability to design multi-agent orchestration workflows (e.g., LangGraph, CrewAI) and use MCPs to connect LLM reasoning to enterprise data actions.
• Proficiency in “vibe coding” and AI-native development workflows using Agentic IDEs and LLM-driven code generation for rapid prototyping and deployment.
• Strong communication skills, ability to influence senior stakeholders, manage ambiguity, prioritise work and balance technical trade-offs.

Additional Details

• Location preference: Bracknell, UK, though other UK locations may be considered.
• Benefits include a generous pension scheme, bonus scheme, private medical and life insurance, up to 31.5 days annual holiday, and training/learning opportunities.