Senior Agentic AI engineer (m/f/d) at Allianz Insurance

Barcelona, Catalonia, Spain -

Full Time

Start Date

Immediate

Expiry Date

05 Jun, 26

Salary

0.0

Posted On

07 Mar, 26

Experience

2 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Agentic AI, Task Decomposition, Planning, Tool-Use, Function-Calling, Memory Management, Multi-Agent Orchestration, Evaluation Design, Feedback Loops, Human-in-the-Loop, API Integration, Latency Optimization, Reliability Optimization, Cost Optimization, Prompt Injection Defenses, Python

Industry

Financial Services

Description

85441 | IT & Tech Engineering | Professional | Non-Executive | Allianz Services | Full-Time | Permanent Allianz Services is proud to serve and be part of Allianz Group, one of the world's leading insurers and asset managers. Our global footprint, with more than 7,300 employees located across ten countries and four continents, enables us to unlock value for our partners across the insurance value chain and deliver superior client experience within Allianz Group. At Allianz Services, we have a strong specialization in core insurance operations, shared services, business analytics, engineering, and management consulting. Our people and our trusted relationships with partners are what matter most. These are rooted in our values of compassion, integrity, and expertise and we strive to continuously evolve, improve, and deliver excellence while contributing successfully to the transformation of Allianz. Caring actively for the environment, for people, and for our customers is what makes us a great place to work and together we are shaping a better and more caring tomorrow. Allianz Services is a company which belongs to the legal entity, Allianz Technology SE. Role overview Are you passionate about AI and eager to see how it transforms the world’s leading insurance brand? Join Allianz Services as a Senior Agentic AI engineer and dive into groundbreaking AI projects within our Data Analytics Solutions team. Collaborate with a dynamic, international team to design, build, and productionize agentic AI systems that deliver measurable business outcomes. Partner closely with the Business Owner to translate goals into working agentic solutions with clear SLAs, governance, and ROI to truly “Make AI Happen.” Apply now and become a key player in shaping the future of AI with us What you do: Co-lead discovery sessions to understand business processes, pain points, and constraints; translate them into solution requirements and success criteria, sharable beyond a technical audience. Architect and implement agentic AI solutions end-to-end: task decomposition, planning, tool-use/function-calling, memory/state management, and multi-agent orchestration. Design evaluators and feedback loops for agents: offline/online evaluation, regression tests, human-in-the-loop review, and continuous improvement. Create robust tool integrations for agents: internal APIs, databases, search/indexing, SaaS systems, and custom actions with strict authorization/scopes. Optimize for latency, reliability, and cost: caching, streaming, batching, retries/backoff, circuit breakers, and budget controls. Implement safety and compliance measures, e.g., prompt injection defenses, audit logging, and least-privilege access for tools. Collaborate cross-functionally with Data Scientists, Platform Engineers, Architects, and Product Managers to convert requirements into shipping agentic capabilities. Write high-quality, testable code and CI/CD for services, actions/tools, and agent state machines; maintain documentation and runbooks We highly welcome candidates with a genuine interest and affinity for Information Technology (IT) and Generative Artificial Intelligence (GenAI), as these attributes are considered valuable assets to our team. What you bring: 3–5 years of relevant professional experience delivering end-to-end data/ML/LLM projects from data preparation to production deployment and monitoring. Strong Python programming and software engineering fundamentals (testing, typing, packaging, async, concurrency). Solid understanding of the machine learning product lifecycle and MLOps practices. Hands-on experience building agentic AI systems in production. Practical knowledge of LLM orchestration frameworks and tools such as one or more of: PydanticAI (preferred), LangChain, LangGraph, AutoGen, CrewAI, or equivalent. CI/CD proficiency and pipeline automation (Jenkins, GitHub Actions), including environment promotion and config/prompt versioning. Experience with containerization and orchestration (Docker, Kubernetes) and service observability (logging, tracing, metrics). Cloud proficiency in at least one major platform; Azure and Databricks preferred, with hands-on experience with MS AI Foundry and MosaicAI. Demonstrable ability to coordinate with Engineers, Architects, and Data Scientists. Ability to self-organize, take ownership, and drive delivery within an agile team. What we offer: We offer a hybrid work model which recognizes the value of striking a balance between in-person collaboration and remote working incl. up to 25 days per year working from abroad We believe in rewarding performance and our compensation and benefits package includes a company bonus scheme, pension, employee shares program and multiple employee discounts (details vary by location) From career development and digital learning programs to international career mobility, we offer lifelong learning for our employees worldwide and an environment where innovation, delivery and empowerment are fostered Flexible working, health and wellbeing offers (including healthcare and parental leave benefits) support to balance family and career and help our people return from career breaks with experience that nothing else can teach

Responsibilities

The role involves co-leading discovery sessions to translate business needs into solution requirements and architecting end-to-end agentic AI solutions covering planning, tool use, and memory management. Responsibilities also include designing evaluation mechanisms, creating robust tool integrations, and optimizing performance for latency, reliability, and cost.