(Senior) AI Engineer (m/f/d)

About the Team

We are a small engineering team inside Natuvion that finds and builds the next high-value additions to Natuvion's product portfolio. We prototype AI-powered solutions for enterprise data transformations — from an idea observed in a customer project to working software.

Our work has three focus areas

• Content corpus: We make Natuvion's transformation knowledge searchable, structured, and available to AI agents. Hundreds of customer projects have produced millions of lines of transformation content. We build the infrastructure that turns this accumulated knowledge into a usable asset.
• Embedded prototyping: We go into customer transformation projects alongside Natuvion's consultants, observe the processes, and identify where AI and technical solutions can accelerate them. Then we build prototypes, test them with real users, and validate the value.
• AI and agentic coding practices: We establish how Natuvion applies AI, both in its products and in its engineering. We develop the methods, tools, and standards the wider organization adopts.

We build on top of the Natuvion Cloud Platform (NCP), using its APIs and shared components as our foundation. The cloud team scales the platform; we focus on what to build on it. The team has strong technical and scientific backgrounds. We care about quality even in prototypes: Disciplined engineering, not vibe-coding. We're part of Natuvion (~500 employees, SAP Gold Partner). That means access to enterprise customers, real transformation projects, and corporate backing but with the autonomy of a small team.

Why join us:

• Real problems, not toy projects. Enterprise data transformations involve millions of records, complex business rules, and tight timelines. The prototypes we build here solve real problems for real customers.
• End-to-end ownership. From observing a process in a customer project to shipping a working prototype. We decide what to build and how.
• AI as a daily tool, not a buzzword. We use agentic coding and AI-assisted development as core working methods. We'll shape how the organization adopts these practices.
• Small team, big reach. A few people with direct access to enterprise customers like BMW, Volkswagen, E.ON, and Lufthansa, a cloud platform to build on, and the freedom to move fast.

Your Role:

• Build and evolve the content corpus infrastructure: graph databases, search indexes, and embedding pipelines.
• Design AI agent capabilities for DCS Pulse: Retrieval, recommendation, and content generation.
• Develop semantic search and retrieval systems across Natuvion's transformation knowledge base.
• Evaluate and integrate new AI models, tools, and techniques as the field moves.
• Establish best practices for AI application and agentic coding across the team and the wider organization.

What you bring:

• Deep experience with AI/ML infrastructure, from classic ML methods to modern LLM-based systems.
• Strong software engineering fundamentals. You ship production-quality code.
• You build systems around LLMs and agentic AI — orchestration, evaluation, guardrails, deployment.
• Familiarity with graph databases, knowledge representation, or search infrastructure.
• You bridge research and production. You read papers and you ship software.
• You care about establishing practices and standards for how AI is applied across an organization.

If you think that you don't meet all the requirements: With us, there are "no limits" to your development! Go your way with us - we support you! Just show us who you are.

How we work:

• Agentic coding as a discipline. We use AI-assisted development tools daily and hold the same quality standards whether code was written by a human or an agent.
• Prototyping with standards. We ship fast, but not at all costs. Prototypes have tests, clean APIs, and documentation. When a prototype proves its value, it should be ready to hand off and not rewritten from scratch.

Our current stack: Python, Type Script, FastAPI, Neo4j, and various LLM APIs. We build on NCP's APIs and shared frontend components. The stack evolves as the problems do. We pick the right tool, not the familiar one.

What you can expect from us:

• An environment that is always open to change and optimization.…