ml engineer for editorial AI applications

Zeit Verlagsgruppe; Berlin; Híbrido; EUR 90 000 - 130 000; Tempo integral

Vantagens oferecidas por esta oferta de emprego

Flexible working hours

Mentoring program

Training opportunities

Travel cost subsidy

Bike leasing

Resumo da oferta

Zeit Verlagsgruppe in Berlin seeks an AI/NLP software engineer to design, develop, and operate AI applications from concept to production, deploying in Docker and Kubernetes, and monitoring performance. You will review AI and human-written code for quality and security, select and optimize language models, and collaborate with editorial, product, and technology teams.

Strong Python skills, ML/NLP expertise, and experience with LLMs/open-source models are essential.

Qualificações

• Degree or equivalent qualifications in CS, computational linguistics, data science, or related field.
• Practical experience in software development and AI/NLP.
• Strong Python skills and experience with ML frameworks.
• Experience with LLMs and open-source models, especially in agent-based systems.
• Proficiency with SQL, shell, and Docker; Kubernetes desirable.
• Ability to review AI- and human-written code for quality, security and performance.
• Experience with code reviews and testing, including automated tests.
• Curiosity, learning mindset, and collaboration with editorial, product and technology teams.

Responsabilidades

• Design, develop, and operate AI and NLP applications from initial ideas through production.
• Extend applications, deploy in Docker and Kubernetes, monitor and resolve issues.
• Review AI and human-written code for correctness, security, maintainability and performance.
• Select, adapt, and optimize language models, especially LLMs and open-source models.
• Analyze data for algorithm development and evaluate AI result quality.
• Collaborate with editorial, product, and tech teams; share AI best practices with others.

Conhecimentos

NLP and ML expertise

Team collaboration

Code review and testing

Analytical and solution oriented

Ability to explain complex topics

Formação académica

Degree in computer science, computational linguistics, data science or related field

Ferramentas

Docker

Kubernetes

Hugging Face / PyTorch / scikit-learn

SQL

Descrição da oferta de emprego

Описание

DIE ZEIT is a leading provider of quality print and online journalism in text and images, data visualizations, and digital storytelling. It offers analytical coverage of world events, opinionated commentary, visual design, podcasts, and video formats.

Задачи

• Design, develop, and operate AI and NLP applications from initial ideas through architecture decisions to production
• Extend applications, deploy them in Docker and Kubernetes environments, monitor them, and analyze and resolve operational errors
• Review human-written and AI-generated code for correctness, security, maintainability, and performance; identify errors, complexity, and security risks
• Select, adapt, and optimize language models, particularly LLMs and open-source models
• Analyze data for algorithm and method development and evaluate the quality of AI results
• Contribute product ideas, follow technical trends in AI, NLP, and AI-assisted software development, and test new approaches with prototypes
• Collaborate with colleagues in editorial, product, and technology teams
• Help other teams integrate AI solutions and share best practices for using AI tools

Требования

• Degree in computer science, computational linguistics, data science, or a comparable field, or equivalent qualifications gained through relevant work experience
• Practical experience in software development and AI/NLP
• Strong practical knowledge of NLP and machine learning, particularly text embeddings, classification, and transformer architectures
• Experience with LLMs and open-source models, particularly in agent-based systems
• Strong Python skills and experience with Hugging Face, scikit-learn, or PyTorch
• Proficiency with SQL databases, shell, and Docker; Kubernetes experience is desirable
• Confident but critical use of AI coding assistants such as GitHub Copilot, Claude Code, or Cursor
• Practical experience with code reviews and manual and automated testing
• Ability to understand and systematically evaluate unfamiliar and AI-generated code, improving its quality, security, maintainability, and test coverage
• Analytical, solution-oriented, and careful approach; critical evaluation of results; responsibility for work quality and security; creativity, flexibility, initiative, curiosity, and willingness to learn
• Interest in the media industry, quality journalism, and the societal impact of AI
• Enjoyment of teamwork, clear communication, ability to explain complex technical topics understandably, and confidence in giving and receiving constructive feedback

Условия

Initially limited to two years 36 Hours per week Flexible working, with part of the week available from home; editorial meetings are hybrid, and the whole editorial team meets once a month Internal and external training opportunities and a mentoring program Onboarding and buddy program Central Berlin office near Potsdamer Platz with public transport connections Travel cost subsidy and bike leasing