Software Engineer — Agentic AI Systems (m/w/d)

As a tech company with locations in Cologne and Osnabrück, we have been specialists in “Interactive Digital Signage” and “Mobile Solutions” at the Point of Sale and Point of Interest for over 15 years, and experts in all projects related to digital transformation.

We are looking for a Software Engineer — Agentic AI Systems (m/f/d) to join our dynamic team at our Cologne location as soon as possible .

WE HAVE EXCITING TASKS FOR YOU

AI Agents & Agentic Systems

• Conceptualization and development of production-ready AI agents and multi-agent workflows for real-world use cases — from role distribution and tool usage to planning and reasoning loops, and memory and state management across multiple agents.
• Selection and integration of agent orchestration frameworks and interop protocols (e.g., LangGraph, CrewAI, AutoGen, Semantic Kernel, PydanticAI, MCP, Vendor Agent SDKs) and their connection to existing REST APIs, event streams, and data pipelines.
• Making and documenting architectural decisions (ADRs) and ensuring non-functional requirements (reliability, cost and latency control, traceability of agent decisions)
• Instrumentation of agent behavior for tracing, evaluation, guardrails, and reproducibility of non-deterministic runs

Data platform & integration

• Extending our event-driven Python data platform (FastAPI services, pipeline orchestration, event bus, relational and vector storage, object storage) to include the interfaces that agents need — new event types, callable tools, retrieval endpoints, human-in-the-loop touchpoints
• Collaboration across disciplinary boundaries — from data engineering to the backend (Python / FastAPI) to DevOps — for the productive operation of agent systems in container deployments (Docker Compose in production; Kubernetes experience is an advantage)

Quality & Operation

• Creation of high-quality, testable code — independently and with AI support (agentic coding tools) — as well as code and pull request reviews.
• Evaluation and monitoring of agent behavior (tracing, guardrails, error handling of non-deterministic behavior)
• Independent problem analysis, troubleshooting and deployments (CI/CD, container and cloud deployments)
• Responsibility for operational safety and stability of the agent systems (troubleshooting in production, release management) — in our ISO 27001 certified environment

Team & Collaboration

• Close collaboration with product owners, stakeholders, and customers to clarify use cases and feasibility.
• Active participation in the agile process (task refinement, effort estimation, sprint planning)
• Knowledge transfer within the team regarding agentic AI practices and tools

WHAT YOU BRING

• Loyalty and integrity
• Enjoyment of teamwork and collaborative work
• Desire for self-determined and self-organized work
• Successfully completed studies in computer science, business informatics or a comparable field of study — or completed training in the relevant field
• Several years of professional experience in software development as well as practical experience with LLM-based applications or agent systems in productive use
• Understanding common LLM application patterns: Prompting, Tool/Function Calling, RAG, Context and Memory Management
• Direct experience with the integration of hosted LLM provider APIs (e.g., Anthropic, OpenAI, Google, Mistral) in production environments — including error handling, rate limit resilience, and cost and latency control.
• Practical experience with at least one agent/orchestration framework (e.g., LangGraph, CrewAI, AutoGen, Semantic Kernel, PydanticAI, Vendor SDKs or comparable in-house developments)
• Practical experience with the Model Context Protocol (MCP) — setting up or using MCP servers to provide agent tools
• Good programming skills in Python (FastAPI or comparable asynchronous web framework); knowledge of TypeScript (Angular / React) is an advantage.
• Solid knowledge of web API design (REST, OpenAPI, versioning, contract discipline)
• Solid data engineering fundamentals (SQL / PostgreSQL, relational data modeling, schema migrations, ETL / ELT concepts)
• Solid knowledge of vector databases and retrieval strategies (pgvector, Qdrant, Weaviate or similar) — central to our retrieval and agent memory layers.
• Practical experience with a pipeline/workflow orchestration tool (Dagster, Airflow, Prefect, Temporal or similar) — the category counts more than the specific product.
• Practical experience with event-driven architectures and stream/queue systems (Redis Streams, Kafka, NATS, RabbitMQ or similar)
• Experience with modern DevOps practices and tools (Git, CI/CD, Docker, ideally Kubernetes)
• Experience with cloud computing providers such as STACKIT, AWS or Microsoft Azure — ideally with deployments under EU data storage
• Awareness of the special characteristics of non-deterministic systems (testing, evaluation, observability of agent behavior)
• Working confidently in an ISO 27001 certified environment — handling secrets, audit trails, client-related data separation
• Excellent comprehension skills, solution-oriented and efficient work style
• Good German and English skills, both spoken and written.

It would also be great if:

• Experience with evaluation/tracing tools for LLM and agent applications (e.g., LangFuse, LangSmith, Braintrust, Weights & Biases)
• Knowledge of accompanying technologies (SQLModel / SQLAlchemy, ORM and database migrations, NoSQL, GraphQL)
• Experience with Identity & Access Management or authentication (OIDC / OAuth2, JWT)
• Familiarity with self-hosted object storage / S3-compatible systems (SeaweedFS, MinIO or similar)
• Prior experience with ingestion pipelines (RSS / Atom, compliance-compliant HTML scraping, email and attachments, document extraction)
• Frontend prototyping with React + Vite + Tailwind or Angular for agent-facing UIs

THIS IS WHAT YOU GET

• Flexible working hours
• Option for hybrid working
• Exciting and challenging projects
• Dynamic and agile working environment
• Attractive salary package
• Space for personal development and growth

So? Did we spark your interest? We'd love to get to know you!

Marie Barn Swallow

bewerbung@newroom-media.de

Job type: Full-time

Salary: €3,500.00 - €5,000.00 per month

Services:

• Corporate events
• Flexible working hours

Application question(s):

• Participation in workshops and project work in the office

Language:

• German (native speaker, C1) (Required)
• English (fluent) (Required)

Work location: On-site