Agentic AI Engineer

As an Agentic AI Engineer embedded within a dedicated Product Squad, you are the technical owner of the end-user facing AI products - specifically, the Generative AI agents and intelligent workflows that deliver tangible business value. Your primary focus is maximising product performance, iterating based on user feedback, and creatively leveraging the capabilities provided by the internal GenAI Platform Team. You will own the entire lifecycle of an AI agent, from prompt design to production deployment and monitoring.

This is a hybrid role, typically working 1 day per week at our Citigen office in London

.
Here's a taste of what you'll be doi

• ngDesign, build, and deploy multi-step, autonomous AI agents and workflows using the standardised frameworks (e.g., LangGraph, custom SDKs) provided by the central GenAI Platform Tea
• m.Master the art of Prompt Engineering to define clear, deterministic, and self-correcting agent behaviour, optimising its reasoning, planning, and task execution for specific product use case
• s.Develop custom tools, APIs, and functions that allow the agent to interact with external enterprise systems (CRM, ERP, databases, etc.) to achieve real-world goal
• s.Leverage and Implement rigorous evaluation and testing pipelines from the GenAI platform(e.g., using RAGAS, synthetic data) to measure agent quality, reliability, factual accuracy, and alignment with user expectation
• s.Dive deep into agent logs, process traces, and memory states to diagnose failure modes (e.g., prompt injection, reasoning loops, hallucinations) and rapidly iterate on agent design and prompts to improve outcome
• s.Work closely with the GenAI Platform Team to identify opportunities for leveraging platform features (like caching, optimised model endpoints) to meet product latency and cost target
• s.Collaborate closely with the Product Manager and UX Designer to translate high-level user needs and business KPIs (e.g., time-to-resolution, conversion rate) into concrete agent behaviours and technical specification
• s.Act as the primary consumer of the internal GenAI Platform, providing critical feedback to the Platform Team on API usability, documentation, and missing capabilitie
• s.Own the agent's deployment process, monitoring, and operational excellence, ensuring the product is secure, stable, and compliant using the governance guardrails established by the Platform Tea

m.
Are we the perfect mat

• ch?2 years of experience in Software Engineering or ML Engineering, ideally with a specialised focus on building Generative AI products or agent-based syste
• ms.Strong proficiency in Python and solid software engineering fundamenta
• ls.Proven, hands-on experience building and deploying complex workflows using modern agentic frameworks (e.g., LangChain, LangGraph, CrewAI, AutoGen, or equivalent internal framework
• s).Practical experience in connecting agents and products to Retrieval-Augmented Generation (RAG) systems, understanding vector stores, chunking strategies, and retrieval performance. You will be connecting to our internal Gen AI Platform's vector databas
• es.Experience working directly with model APIs (e.g., OpenAI, Anthropic, Gemini, or internal endpoints) and managing model configurati
• on.Familiarity with the production deployment lifecycle, including Git, CI/CD pipelines, containerisation (Docker), and consuming microservices via REST/gR
• PC.A strong passion for understanding end-user problems and designing AI solutions that are intuitive, reliable, and delightful to u
• se.The ability to understand the business context and define the success of an agent not just technically, but by its measurable impact on key business metri
• cs.Excellent analytical skills to debug complex, non-deterministic systems and iteratively drive towards predictable, high-quality outp
• ut.Effective collaboration skills to work across a cross-functional squad (Product, UX, Backend) and ability to clearly communicate technical tradeoffs and capabiliti