About the job

I'm partnered exclusively with a global energy company (6,000+ employees, 20 countries, 10+ million homes served) building out its AI engineering capability from its Berlin tech hub. The team ships production NLP and conversational AI at real scale: RAG-based knowledge systems and agentic voicebot products used by millions of households.

• What you'll do:Build and evolve RAG pipelines: retrieval, chunking, embeddings, vector databases
• Design agentic systems and conversational workflows (graph design in LangGraph)
• Fine-tune models and develop custom classifiers
• Build evaluation pipelines: metrics, benchmarking, quality assessment of language models
• Monitor production models including drift detection and log analysis
• Prompt engineering alongside business experts, translating requirements into AI solutions
• Mentor junior engineers and shape technical standards

• Your experience:Degree in Computer Science, Data Science, Mathematics or comparable
• 3+ years in Data Science, ML or AI Engineering
• Strong Python and solid ML/NLP fundamentals: model training, statistical analysis, evaluation methods
• Hands-on with the modern LLM ecosystem: PyTorch, TensorFlow, HuggingFace, LangChain, LangGraph
• RAG architectures, vector databases, orchestration frameworks
• Cloud experience, ideally Azure
• Voicebot or conversational AI experience is a plus
• Fluent German (C1 minimum) and very good English

• Details€85,000 - €110,000
• 30 days holiday per year
• Hybrid working with up to 50% remote
• Flexible working hours
• Strong learning and development culture with ongoing training opportunities
• Agile working environment within a globally established business

Interested? Reach out directly to Norman Cistovas at Xcede or apply below.

·        Generative AI and LLM Engineering: LLM Application Development, Retrieval Augmented Generation (RAG), Prompt Engineering, LangChain, LangGraph, AI Agents, Multi Agent Systems, Model Context Protocol (MCP), Hugging Face Transformers, OpenAI GPT, Anthropic Claude, Google Gemini, Meta Llama, Groq

·        Vector Search and Data Engineering: ChromaDB, FAISS, Pinecone, Sentence Transformers, Semantic Search, Python, SQL, PostgreSQL, Pydantic

·        Machine Learning and Deployment: Deep Learning, TensorFlow, PyTorch, Natural Language Processing, Computer Vision, FastAPI, SQLAlchemy, Alembic, Docker, Docker Compose, MLOps, Microsoft Azure, CI/CD

·        Analytics, Tools and Methods: Tableau, Power BI (DAX), Git, GitHub, JWT Authentication, REST API Design, Agile Delivery