Senior AI Engineer

Senior AI Engineer Location: Central London Working pattern: Hybrid, ideally 2–3 days per week onsite Salary: £90,000 Employment type: Permanent About the Company Our client is a SaaS company based in central London, building AI-enabled products used by business customers at scale. They currently have a customer base of around 2 million users. They are investing heavily in production-grade AI systems and are looking for a Senior AI Engineer who can help design, build, and operate intelligent agentic systems in a real-world SaaS environment. This is not a role focused on simple chatbot prototypes or basic RAG demos. The team is looking for someone who has shipped production AI systems and can explain how they were architected, orchestrated, monitored and made resilient. The Role As a Senior AI Engineer, you will work on the design and delivery of production-ready AI systems, with a particular focus on agentic workflows, retrieval-augmented generation, AWS-native deployment and robust orchestration. You will be expected to contribute across architecture, implementation, evaluation, and production operations. The successful candidate will have strong hands-on engineering skills and proven experience building AI systems that go beyond standard retrieval pipelines. A strong candidate will be able to describe a named production agentic system they have shipped, including the orchestration approach, how tools or agents interacted, how state was managed and how failure modes were handled. Key Responsibilities * Design, build, and maintain production-grade AI and agentic systems within a SaaS product environment. * Develop agentic workflows using LangGraph or equivalent orchestration frameworks. * Build and improve RAG and retrieval pipelines, including ingestion, chunking, embedding, search, reranking, evaluation, and monitoring. * Design robust orchestration patterns for AI systems, including tool use, routing, state management, retries, fallbacks, human-in-the-loop processes, and error handling. * Deploy and operate AI systems in AWS-native environments, ideally using services such as Amazon Bedrock, Lambda, ECS, EKS, Step Functions, OpenSearch, CloudWatch, and related tooling. * Work closely with product, engineering, and leadership teams to turn business problems into reliable AI-powered product capabilities. * Evaluate model and system performance using appropriate metrics, testing approaches, and feedback loops. * Use AI coding assistants as part of your daily engineering workflow to improve productivity, code quality, testing, and delivery speed. * Contribute to engineering best practices around security, observability, scalability, reliability, and maintainability. Required Experience * Strong software engineering background, ideally in Python and modern backend engineering environments. * Proven experience shipping production AI, LLM, or agentic systems. * Hands-on experience with LangGraph or an equivalent framework for agent orchestration. * Deep understanding of RAG and retrieval pipelines, including vector search, embeddings, document processing, query transformation, reranking, and evaluation. * Experience designing AI systems that handle real-world failure modes, such as tool failures, hallucinations, poor retrieval results, timeouts, partial outputs, and low-confidence responses. * Strong AWS experience, with a preference for candidates who have worked with Amazon Bedrock in production or near-production environments. * Experience building observable, testable, and maintainable AI systems. * Daily or frequent use of AI coding assistants such as Cursor, GitHub Copilot, Claude Code, or similar tools as part of an engineering workflow. * Ability to communicate technical decisions clearly and explain architectural trade-offs to both technical and non-technical stakeholders. Desirable Experience * Production experience with Amazon Bedrock, Bedrock Agents, Bedrock Knowledge Bases, or Bedrock Guardrails. * Experience with LangChain, LlamaIndex, Semantic Kernel, AutoGen, CrewAI, Haystack, or custom agent orchestration frameworks. * Experience with AWS OpenSearch, Step Functions, Lambda, ECS, EKS, SageMaker, CloudWatch, IAM, and CI/CD tooling. * Experience with LLM evaluation, guardrails, monitoring, prompt/version management, and AI observability tools. * Experience working in a SaaS product environment. * Experience with multi-agent systems, tool-calling workflows, workflow graphs, or state-machine-based AI applications. What a Strong Application Looks Like * The strongest candidates will be able to provide evidence of a production system they have personally helped design and ship. This should include: * A named AI or agentic system they worked on. * The business or product problem it solved. * The orchestration approach used, for example LangGraph, state machines, tool routing, workflow graphs, or custom orchestration. * How retrieval and RAG were implemented. * How the system handled failure modes, retries, fallbacks, poor retrieval, tool errors, hallucinations, or low-confidence outputs. * How the system was deployed and operated, particularly within AWS. * How AI coding assistants are used in their day-to-day engineering workflow. Working Pattern The company is based in central London and operates a hybrid working model. The preference is for someone who can work onsite 2–3 days per week, collaborating closely with the engineering and product teams. Ideal Candidate Summary This role would suit a Senior AI Engineer, Applied AI Engineer, AI Platform Engineer, Senior Backend Engineer with strong GenAI experience, or ML/LLM Engineer who has built production-grade agentic systems. The ideal candidate will combine strong software engineering fundamentals with hands-on experience in agent orchestration, AWS-native deployment, RAG pipelines, and practical day-to-day use of AI engineering tools.