We Are Waiting for You We are looking for a Senior AI Engineer - Agentic Workflows to develop and deploy generative AI systems at Zurich Technology Delivery Center.

Who Are

We?

Zurich Insurance

Group’s Technology Delivery Center (TDC) in Barcelona supports global business with software development, data analytics, cybersecurity, and IT infrastructure. Why Join? Work on cutting‑edge generative AI systems at scale Collaborate with a specialized AI team including Data Scientists, AI Engineers, and ML Optimization specialists Build foundational patterns and frameworks that shape our approach to agentic AI What Can You Expect?

At the core, you will design and implement agentic workflows that translate business problems into production‑ready solutions, working closely with product and business stakeholders.

Responsibilities Design Agentic Workflows: Architect end‑to‑end workflows using Lang Graph or similar, decomposing complex problems into composable steps, states, and policies.

Problem Decomposition & Solution Design: Translate high‑level business problems into structured agentic solutions and evaluate design options.

Build

Skills, Tools, and Agents: Implement reusable components and compose them into maintainable workflows.

Choose Building Blocks: Decide when to build a skill, tool, agent, or workflow.

Model Strategy & Integration: Integrate multiple LLM providers, optimising for latency, cost, and quality.

Performance Optimization: Optimize prompts, token usage, chunking, and inference efficiency.

Evaluations & Quality: Define offline and online evaluation strategies, datasets, metrics, and guardrails to validate behaviour and track regressions.

Data Structuring & Validation: Design typed schemas for data flows, tool contracts, and structured outputs.

Observability & Debugging: Implement logging, tracing, and monitoring using tools like Langfuse or Lang Smith.



AI‑Accelerated Engineering: Use AI assistants (Claude Code) for prototyping, refactoring, testing, and documentation. Cross‑Functional Collaboration: Partner with data scientists, engineers, and product to translate requirements into robust capabilities.

Qualifications Must have: 5+ years of Python development with strong software engineering fundamentals. Hands‑on experience building production LLM‑based systems. High ownership and proactive communication. Proficiency with Lang Graph or similar agentic frameworks (Lang Chain acceptable). Strong evaluation methodology skills for LLM/agent quality. Ability to decide when to build a skill vs tool vs agent vs workflow.

Experience decomposing complex, ambiguous business problems into scalable solutions.

Nice to have: Comfort using AI assistants (e.g., Claude Code) in daily engineering. Familiarity with LLM observability tools (Langfuse, Lang Smith).

Background with vector databases and RAG patterns. Knowledge of cost optimisation and token accounting in LLM systems.

Async

Python production experience (asyncio, concurrent handling). Azure deployment and cloud‑native architecture experience. Understanding of AI output quality metrics and evaluation frameworks.

Benefits Competitive salary and annual bonus. Adaptable working model and up to 25 days of abroad work per year. Home office setup support and monthly allowance. Extensive internal and external training opportunities. Ticket restaurant, health, life, accident, and retirement insurance. Referral bonus and exclusive employee discounts. Stock options and mortgage benefits.

Location

Barcelona, Poblenou.

Work arrangement Hybrid remote working; full time schedule. At Zurich, we are an equal opportunity employer. We attract and retain the best-qualified individuals available, regardless of race/ethnicity, religion, gender, sexual orientation, age, or disability.

📌 Senior ai agent & workflow engineer (Madrid)
🏢 Tamarind Intelligence
📍 Madrid