Lead / Principal AI Engineer

Lead / Principal AI Engineer at AKQA in London shapes production-grade AI systems for digital experiences. You’ll design, build and uplift intelligent workflows, agentic solutions and scalable architectures in collaboration with cross-functional teams. This role blends deep engineering with technical leadership, balancing innovation with reliability and cost. You’ll translate ambitious ideas into production-ready technology that delivers meaningful business value for top brands.

• Architect, develop and deploy production-ready AI systems including agentic workflows and multi-agent orchestration
• Lead the technical design and implementation of AI-powered digital products and platforms to ensure scalability, security and maintainability
• Design and integrate agent tools, APIs and MCP interfaces for secure enterprise integration
• Develop hybrid AI architectures blending LLMs with specialised models and retrieval techniques to improve cost and performance
• Build evaluation frameworks, automated testing, guardrails and observability for model behaviour and operational performance
• Apply model selection, prompt engineering, structured outputs, and retrieval strategies balancing technical and commercial goals
• Design and develop backend services, APIs and integration layers using Python, TypeScript/Node.js and cloud technologies
• Collaborate with frontend, mobile and product teams to integrate AI into web apps (React, Next.js)
• Implement token management, caching, latency control, fallback mechanisms and graceful degradation for AI services
• Establish CI/CD for AI systems with automated evaluations, regression tests and reliable deployments
• Engage with cloud/model platforms (Azure OpenAI, AWS Bedrock, Vertex AI, OpenAI, Anthropic) to deliver enterprise-ready solutions
• Partner with creative, strategy, UX and tech teams to explore innovative AI applications
• Provide technical leadership and architectural guidance to engineering teams
• Communicate technical approaches to clients and stakeholders to shape value-driven solutions
• Own technical architecture while contributing hands-on engineering and delivery
• Work in a digital agency context, collaborating across multidisciplinary teams
• Exposure to fine-tuning smaller models (e.g., Llama, Mistral) for domain-specific apps would be advantageous
• Experience with design-to-code workflows, Figma integrations, design systems or generative UIs would be beneficial
• Hybrid work environment with in-office days; focus on collaboration, innovation and client impact

• Backend, platform or software engineering with distributed systems and cloud experience
• Hands-on production experience with generative AI or agentic systems
• Advanced Python and/or TypeScript/Node.js proficiency
• Strong understanding of LLMs, agentic workflows, RAG, embeddings, vector DBs and retrieval techniques
• Experience with agent orchestration frameworks (LangGraph, CrewAI, AutoGen, Vercel AI SDK or equivalents)
• Experience building AI evaluation and observability capabilities (LangSmith, Arize Phoenix, Braintrust, Helicone, OpenTelemetry)
• Solid software architecture, security, data management and enterprise integration knowledge
• Experience with automated testing and CI/CD for non-deterministic AI systems
• Understanding of cloud infrastructure and AI deployment/scaling considerations
• Awareness of modern digital product development (web apps, headless CMS, APIs, frontend tech)
• Ability to balance ambition with cost, latency, performance and maintainability
• Curiosity and pragmatism in adopting new tech, with focus on real value from AI
• Strong communication of complex concepts to clients and stakeholders
• Willingness to own architecture while remaining hands-on
• Experience in a digital agency, technology consultancy or product-led environment
• Familiarity with fine-tuning open-weight models for domain use would be advantageous
• Experience with design-to-code tools and generative UI is beneficial
• Collaborative, proactive and technically curious mindset

• collaborative
• strong communication
• problem-solving
• Python
• TypeScript/Node.js
• LLM architectures