AI / ML Engineer - Agentic AI ( m/f/d)

Rohde & Schwarz; Dresden; On-site; EUR 68,000 - 95,000; Full time

Job summary

Rohde & Schwarz in Dresden is seeking an expert to design agent-based AI systems from concept to production. You will craft agent architectures, model harnesses, prompt strategies, and robust context management for scalable AI solutions.

You will implement retrieval-augmented generation, vector search, and multi-stage orchestration with tool calling and MCP integrations, while evaluating safety, monitoring and failure modes, and building reproducible evaluation pipelines for production.

Responsibilities

• You design agent architectures, model harnesses, prompt strategies, and context management for agent-based AI systems.
• You implement retrieval-augmented generation, vector search, and multi-stage orchestration with tool calling and MCP integrations.
• You evaluate generative behavior, security, and failure modes such as hallucinations, and establish monitoring and countermeasures.
• You build reproducible evaluation pipelines and conduct metric-driven experiments.
• You transition solutions into production using MLOps and software engineering practices while remaining closely aligned with users and stakeholders.
• You stay up to date with the state of the art in foundation models and agentic AI and evaluate their potential for use on the platform.

Job description

Introduction

You design agent-based AI systems that make a real difference from the idea stage to actual implementation.

Tasks

• You design agent architectures, model harnesses, prompt strategies, and context management for agent-based AI systems.
• You implement retrieval-augmented generation, vector search, and multi-stage orchestration with tool calling and MCP integrations.
• You evaluate generative behavior, security, and failure modes such as hallucinations, and establish monitoring and countermeasures.
• You build reproducible evaluation pipelines and conduct metric-driven experiments.
• You transition solutions into production using MLOps and software engineering practices while remaining closely aligned with users and stakeholders.
• You stay up to date with the state of the art in foundation models and agentic AI and evaluate their potential for use on the platform.