Software Developer – RAG, Knowledgraph & Agentic Systems (m/f/d)

Company description Every story counts. Design your own – together with us at Riverty. Where does that lead you? In one of our 30 hybrid working environments – designed for the exchange of ideas, common learning and the co-design of our work. Be part of an international community with over 4,000 people from almost 80 nations in 12 countries. Set up by a joint mission: with empathy, modern technology and data-based insights to ensure that people and businesses remain in the river. With payment processes that are done for them – so You don't have to worry about it. And this is not all: we are part of the Bertelsmann group, which is in family ownership. Etablated. Entrepreneur. In a dynamic industry. We enable flexible payment options in various industries, simplify the financial management of well-known brands and help people reduce debt and gain financial security. In short, we design FinTech. Description We are looking for Software Developer – RAG, Knowledgraph & Agentische Systeme (m/w/d) Join our team at one of our locations in Berlin, Baden-Baden or Verl – hybrid working conditions available.

Your mission You work closely with our product and specialist teams and develop AI-supported functions that come directly with our users. Your focus is on RAG pipelines, knowledgraph, context engineering and agentic multi-step workflows. You translate the possibilities of modern AI into practical, reliable and scalable product functions. In this role, you play a central role in building and developing Riverty's AI platform to sustainably increase the productivity of our software development. You design and implement AI-based solutions based on Large Language Models, Agent Architectures and Knowledgraph, creating the basis for automation and intelligent developer workflows throughout our technology organization. Your tasks Development and optimization of RAG pipelines, including retriever components, embeddings, indexing processes and evaluation logics. Integration and use of Knowledgraph to provide AI systems and agents with structured context. Implementation of agentic multi-step workflows with MCP clients, orchestration logic and supporting tools. Development of prompting strategies, Chunking concepts and methods for context preparation, adapted to specific product requirements. Integration of AI models into existing platforms and systems of Riverty. Performance optimization, benchmarking and cost optimization for RAG and agentic architectures. Close cooperation with Platform Engineers in the use of common SDKs, gateway batteries and architecture standards. Creating and maintaining a clear technical documentation and an active contribution to best practices in context engineering.

Qualifications Your profile Found experience in software development with Java, Python or TypeScript (at least 4 years). Experience with or pronounced motivation for working with RAG pipelines, retrieval systems, vector databases or graph technologies. Ability to translate AI potentials into concrete, user-oriented product functions. Structured, reliable operation with high self-responsibility and strong implementation focus. Strong interest in data-driven systems, search technologies and context architectures. Team-oriented working and clear and effective communication in in interdisciplinary teams. German and English knowledge in word and writing (C1+)