Generative AI Specialist

Are you excited by the move from chatbots to reliable, production-grade AI agents? Do you enjoy turning emerging AI capabilities into customer experiences that are useful, safe, measurable, and genuinely scalable? Join the Generators squad in If’s Oasis tribe and help shape the next generation of digital customer experiences in insurance. Our GenAI-powered assistant is live across Denmark, Sweden, and Norway, serving customers on web and app. Now we are expanding across the Nordics and evolving the assistant from a strong conversational experience into a more personalized, connected, and action-oriented AI product. You will be central to that journey, working on real production systems with a modern stack (LangGraph, Langfuse, Azure): designing, building, evaluating, and operating AI systems that can use internal knowledge and APIs responsibly to help customers get the right guidance at the right moment. About the role This is one open position, and we are open to both junior and senior candidates. As our Generative AI Specialist, you will be a hands-on engineer in the Generators squad, building production AI systems alongside the team, from agent design and retrieval to evaluation and monitoring. Your responsibilities include: Designing and building production-grade agentic workflows with tool and API use, orchestration, guardrails, and safe handover paths. Building the agent harness: the runtime that makes agents dependable in production, including context management, memory, and observability. Scaling our AI assistant across Nordic markets. Integrating AI systems with internal APIs and data to deliver personalized support. Improving retrieval, grounding, and response quality through systematic experimentation. Establishing evaluation practices (test sets, regression testing, human and safety evaluation) and monitoring quality, latency, and cost. Building improvement loops that use evaluations, customer feedback, and production signals. Collaborating with the Product Owner and the team through code reviews and knowledge sharing. We have a hybrid setup with fixed team days onsite. Occasional travel to Nordic markets and our Kuala Lumpur office is part of the role. We offer Here are some of the benefits of working at If: An inclusive work environment where everyone is welcome Career and development opportunities in the biggest insurance company in the Nordics Social activities and a highly skilled professional environment Possibility of a hybrid workplace A health-promoting workplace with e.g., fitness programs and various sports activities Attractive insurance benefits About the team You will be a hands-on squad member in Generators, an agile team that builds customer-facing solutions driven by real user insights. We work with a clear build, measure, learn mindset and continuously validate what creates the most value for our customers. Collaboration, code quality, and a team-first mindset matter deeply to us. You will work closely with colleagues across engineering, product, and business to turn ambitious AI opportunities into reliable customer experiences. Junior colleagues are supported by experienced teammates through pairing and code reviews, and seniors help shape the team’s direction. Who are you? You are an engineer, at junior or senior level, who is curious about AI and enjoys building, testing, and improving systems in production. You are a team player who is pragmatic, unafraid to challenge assumptions, and thrives in a cross-Nordic, cross-functional, hybrid environment. Must have (all levels): A university degree in a technical or quantitative field, such as computer science, engineering, mathematics, mathematical economics, physics, or statistics. Hands-on Python experience, ideally with FastAPI, Pydantic, and pytest. Practical experience building LLM-powered or agentic applications, ideally with LangGraph or LangChain, and sound knowledge of agentic best practices: tool use, orchestration, state management, guardrails, human-in-the-loop, and safe fallbacks. Experience with AI evaluations (test sets, regression testing, LLM-as-judge) and observability of LLM applications (we use Langfuse). Working knowledge of Git, CI/CD (Azure DevOps), Docker, and cloud, ideally Azure. The ability to explain technical trade-offs clearly to engineers and business stakeholders. Senior level / nice to have: Proven experience running LLM-powered or agentic systems in production, including cost, latency, and incident management. Knowledge of RAG: embeddings, vector search (we use Qdrant), reranking, and grounding. Experience designing an agent harness or comparable runtime: context management, memory, access control, and safe deployment. Azure (OpenAI or AI Foundry, Key Vault, App Configuration), infrastructure as code with Bicep, and secure API design. Data and platform services such as Redis, MongoDB, and Qdrant, and tracing with Langfuse, LangSmith, or OpenTelemetry. Mentoring colleagues and shaping technical di