platform engineer

Enfint; Berlin; Vor Ort; EUR 90.000 - 140.000; Vollzeit

Benefits dieser Stelle

Flexible working hours

Up to 3 days mobile work per week

Workation in Europe (EEA)

Health and wellbeing offers via evermo

Childcare subsidies

Canteen with organic, regional food

Professional development opportunities

Zusammenfassung

The Cornelsen Group in Berlin is seeking an experienced professional to operate and advance a LiteLLM-based LLM proxy, providing reliable access to Azure OpenAI and AWS Bedrock for product teams.

You will define SLOs, handle failover, lead incident response, ensure GDPR-compliant routing and EU data residency, and advise teams on model choices, documentation, and observability using Langfuse.

Qualifikationen

• Experience operating an LLM proxy or gateway in production (prefer LiteLLM).
• Strong Kubernetes knowledge and familiarity with at least one cloud platform (Azure, AWS, or GCP).
• Experience with model providers such as Azure OpenAI or AWS Bedrock.
• Clear ability to influence others without formal authority.
• Ability to explain technical trade-offs to product teams and non-technical stakeholders.
• Excellent documentation and knowledge-sharing skills.

Aufgaben

• Operate and continuously develop the LiteLLM-based LLM proxy to provide product teams with reliable access to LLM providers like Azure OpenAI and AWS Bedrock.
• Manage the gateway as a production-critical service by defining SLOs, provider failover, and rate-limiting.
• Lead incident response for the gateway service.
• Ensure data protection compliance, including GDPR-compliant routing, EU data residency, and adherence to the EU AI Act.
• Advise product teams on model selection, integration patterns, and prompt strategies.
• Create clear developer documentation and best-practice examples.
• Support teams with onboarding and daily platform usage.
• Act as the primary point of contact for product teams working with LLMs.
• Establish centralized observability for LLM and agent applications using tools like Langfuse.

Kenntnisse

Kubernetes

Cloud platforms (Azure/AWS/GCP)

Stakeholder communication

Documentation

Observability

German & English bilingual

Consulting to product teams

Tools

Langfuse

Azure OpenAI

AWS Bedrock

Jobbeschreibung

Описание: The Cornelsen Group is a leading provider in the German-speaking education market, offering innovative cross-media educational solutions for all stages of life. The organization focuses on creating meaningful educational content with high social relevance.

Tasks

• Operate and continuously develop the LiteLLM-based LLM proxy to provide product teams with reliable access to LLM providers like Azure OpenAI and AWS Bedrock.
• Manage the gateway as a production-critical service by defining SLOs, provider failover, and rate-limiting.
• Lead incident response for the gateway service.
• Ensure data protection compliance, including GDPR-compliant routing, EU data residency, and adherence to the EU AI Act.
• Advise product teams on model selection, integration patterns, and prompt strategies.
• Create clear developer documentation and best-practice examples.
• Support teams with onboarding and daily platform usage.
• Act as the primary point of contact for product teams working with LLMs.
• Establish centralized observability for LLM and agent applications using tools like Langfuse.

Requirements

• Practical experience in operating an LLM proxy or gateway in production, preferably with LiteLLM.
• Solid Kubernetes knowledge and familiarity with at least one cloud platform (Azure, AWS, or GCP).
• Experience with model providers such as Azure OpenAI or AWS Bedrock.
• Strong consulting mindset with the ability to influence others without formal authority.
• Ability to communicate technical trade-offs clearly to product teams and non-technical stakeholders.
• Excellent documentation and knowledge-sharing skills.
• Experience with observability tooling, ideally Langfuse or similar stacks.
• Fluent in German and English.
• Nice to have: Experience in educational technology or similar regulated domains.

Conditions

• 30 days of vacation and flexible working hours.
• Up to 3 days of mobile work per week.
• Up to 20 days of workation in Europe (EEA) per year.
• Access to health and well-being offers via evermood and Wellhub.
• Childcare subsidies, book discounts, company pension scheme, and capital-forming benefits.
• Sustainable mobility options including company bike leasing and reduced Deutschlandticket.
• On-site canteen with organic, regional food.
• Professional development opportunities including training, conferences, and e-learning.
• Regular team events and sports activities.