Machine Learning Engineer – LLM & Agentic AI

Project overview

For our client, we are looking for an experienced Machine Learning Engineer – LLM & Agentic AI.

Responsibilities

Knowledge

We look forward to receiving your CV with details on relevant project experience, hourly rate, and possible start date.

We are happy to assist you with any questions.

• Implementation and integration of LLM-based dialog systems and semantic search solutions
• Setup and maintenance of RAG systems, including knowledge base management, indexing and updating of structured and unstructured data sources, and source-separated storage
• Conception and further development of graph-based agentic dialog systems from intent recognition through knowledge retrieval and extraction to deterministic dialog management and connection of business services
• Go-live of agentic systems, including tool and function calling, and integration into existing portal systems
• Responsibility for operation, deployment, and monitoring of AI solutions and infrastructure within the scope of MLOps
• Ensuring observability through structured logging, error analysis, and dashboards and alerting (ideally Grafana)
• Optimization of response quality, latency, and technological basis based on monitoring data, test results, and user feedback
• Setup of systematic quality measurement with versioned eval sets, metric catalogs, and automated evaluation runs to detect quality regressions
• Development of data pipelines for data preparation and model creation, as well as testing new AI application fields up to go-live
• Security through guardrails, protection measures against jailbreak and prompt injection attacks, and filtering of inputs and outputs
• Analysis of penetration tests and implementation of security findings, as well as data protection compliant design, especially when handling personal data in logging and observability processes
• Deployment and operation in production-like environments with Kubernetes or OpenShift, ArgoCD, and Jenkins
• Completed degree in Computer Science, Business Informatics, or comparable qualification
• At least 3 years of professional experience as a Machine Learning Engineer with practical experience in development, implementation, and optimization of scalable ML solutions
• At least 2 years of experience in agile development teams
• German language skills at least B2 (language certificate or native speaker)
• Experience in securing LLM applications with guardrails, protection measures against jailbreak and prompt injection attacks, and input/output filters
• Experience with agentic systems, graph-based architectures, tool and function calling, and deterministic dialog management
• Knowledge in vector search and semantic indexing, as well as practical experience with Milvus (setup, loading, and recovery of searchable knowledge bases)
• Experience in backend development with FastAPI
• Knowledge of LLM orchestration frameworks such as LangChain and LangGraph
• Experience in productive AI operation (MLOps), especially monitoring, logging, observability, error analysis, and deployment
• Experience with dashboards and alerting, ideally Grafana
• Experience in quality measurement of GenAI solutions with versioned eval sets, metrics, automated evaluation runs, and detection of quality regressions
• Experience with data protection compliant design of AI systems, especially when handling personal data in logging and observability processes
• Good knowledge of Kubernetes or OpenShift, ArgoCD, and Jenkins
• Experience in agile delivery structures, ideally SAFe, including participation in a SAFe train and PI planning
• Ideally, project experience in public administration, regulated environments, or statutory social insurance carriers

Project details

16.11.2026

until 01.12.2027

90% remote, 10% on-site in Hanover, Hamburg, or Munich

Utilization: Full-time

Goal

Development and further development of scalable AI and GenAI solutions and their integration into existing business applications. Focus areas include LLM-based dialog systems, RAG systems, agentic architectures, and their secure and high-quality go-live and operation.

• Start: flexible
• Duration: 12 Monate
• Remote: hybrid
• Location: Hannover, Deutschland, DE
• Languages: German