Principal ML Engineer (Agentic AI)

TL;DR

Principal ML Engineer (Agentic AI): Designing autonomous agents that generate prompts, build tools for real-world actions, and validate outputs using LLMs, RAG, and agent architectures with an accent on end-to-end ML systems, scalable data pipelines, and production deployment. Focus on architecting reliable infrastructure, integrating with APIs, establishing observability frameworks, and mentoring engineers.

Hybrid in Berlin, Germany: 2 days a week in Berlin campus required. Relocation support to Berlin available.

Company

World’s pioneering local delivery platform operating in 65+ countries, headquartered in Berlin, Germany, and listed on the Frankfurt Stock Exchange.

What you will do

• Design and own end-to-end ML and data systems from ingestion to production deployment
• Architect scalable data pipelines for RAG, embeddings, and real-time processing
• Build and operate production-grade ML services, APIs, and infrastructure with IaC, containerization, and orchestration
• Integrate ML systems with external APIs, tools, and platforms for automation
• Establish monitoring, evaluation, and observability across data, models, and systems
• Mentor engineers and set technical direction for the team

Requirements

• Strong experience designing and scaling production-grade ML systems and data platforms
• Deep expertise in data engineering, ML pipelines, RAG, and embedding workflows
• Proven experience building reliable data infrastructure with data quality guarantees
• Strong Python and SQL skills, Docker, Kubernetes, cloud environments
• Experience with IaC (Terraform) and reproducible scalable systems
• Hands-on integration of ML with APIs/services and production operations with monitoring

Nice to have

• Experience with LLMs, agent architectures (LangGraph, AutoGen, CrewAI)
• Familiarity with synthetic data, evaluation systems, AI feedback loops
• Agent/ML observability tools (Langfuse, LangSmith)
• Model serving, routing, inference optimization
• Open-source models (LLaMA, Mistral) or custom inference
• System reliability and safety patterns in AI

Culture & Benefits

• Hybrid model with 2 days/week in Berlin campus
• 27 days holiday + extra days after 2-3 years
• €1,000 educational budget, language courses, Udemy, parental support
• Health checkups, meditation, gym/bicycle subsidy
• Employee share plan, sabbatical bank, transport discounts, insurance, pension
• Meal vouchers, food perks, corporate discounts