We are looking for an AI Agent Engineer to translate identified AI and automation opportunities into working, production-grade solutions. Use cases have been defined at a high level; the next step is to build them — which requires deeper understanding of each process than what is captured today.This is a stakeholder-facing engineering role. You will spend time directly with the colleagues who execute the work — claims handlers, underwriters, operations specialists, back-office teams — because that is where the real specification lives. The quality of those engagements will directly shape the quality of what you ship.Contract duration: PermanentSchedule: Full-time (8 hrs/day, Monday to Friday)Your responsibilitiesRun structured working sessions with end users to understand how the work is actually performed today, including exceptions, manual workarounds, and tacit decision rulesTranslate that understanding into engineerable specifications — process steps, decision criteria, data sources, edge cases, success metricsBuild prototypes and proofs of concept, demo them back to the same end users, iterateImplement AI use cases such as: o Data extraction from documents (PDFs) o Automated form filling o Case classification and routing o Decision support o Document summarization and transformationIntegrate AI solutions with internal systems via APIsBring successful prototypes to production — scalable, observable, continuously evaluated, and maintainable.Required skillsEliciting detailed requirements directly from end users through structured working sessions, and documenting outcomes in a form engineering teams can build uponPrompt, Context, and Harness EngineeringMCP / Agent-to-Agent conceptsPython and microservicesCloud (AWS / Azure)API integration experienceSoftware engineering fundamentals and TDDEvaluation discipline for non-deterministic systems — golden datasets, regression on prompt changes, observability for latency, cost, and qualityFluent in English (written and spoken).How You WorkPragmatic and outcome-focused — quick wins over theoryTreat time spent with end users as engineering work, not overheadFluent across engineering, business, and end-user contexts; can hold technical depth without losing the user's perspectiveClear, respectful communicator; able to explain technical concepts to non-technical end users and to incorporate their feedback constructivelyStrong ownership combined with an experimentation mindsetWork within regulated-data constraints — GDPR, PII handling, model governance, no-train clauses.Desirable skills (Nice to have)Knowledge of the Insurance domainGerman language skills.Compensation & benefitsVery stable work environment, as part of a large Insurance Multinational GroupFlexible work program, respecting your own private timeAttractive remuneration packageLunch ticketsSpecial Days vouchers (Women’s Day, Children's Day, Christmas, Easter)Private pension and private health insurance contributionTime off when holidays fall on a weekend (up to 3 days/year)Exchange of experience and training with international professionals, as a premise for personal developmentTeam building events. Show more Show less