Working Student AI Engineer – LLMs, RAG & AI Agents

IMPORTANT Application note Please apply only through this job portal. Kindly do not send applications or recruitment messages by email or LinkedIn. Only applications submitted through the portal will be considered. Thank you. About AgentixLake AgentixLake GmbH is a Data and AI engineering consultancy based in Hamburg, Germany. We design, build and operate production-grade data platforms and AI systems in our customers’ cloud accounts or on-premises environments. We are an AWS Services Partner and a Certified Services Partner in Anthropic’s Claude Partner Network. Our commercial focus is DACH and the Gulf region, alongside selected international opportunities. Your role Build AI application components that are ready to be tested, deployed and maintained. You will take ownership of defined engineering work within our generative AI projects, collaborating with colleagues on implementation, review and operational quality. Our team builds enterprise AI solutions that connect models with useful data and tools. Your work may contribute to document intelligence, knowledge assistants, governed analytics or AI agents that support engineering and operational workflows. What you will do Develop Python services and integrations connecting language models to enterprise data, APIs and tools. Implement document ingestion, chunking, embeddings and retrieval, with source references that users can inspect. Build structured extraction and validation workflows for complex documents and downstream business systems. Implement defined agent steps, including tool calls, permission checks, error handling and human approval. Maintain evaluation cases and compare retrieval quality, answer usefulness, latency and cost as the application changes. Contribute automated tests, deployment configuration, logging and documentation through reviewed pull requests. What you bring Current university enrolment in computer science, AI, data science or a related technical discipline. Strong practical Python skills and evidence of software you have delivered for actual users, including responsibility for testing and deployment. Hands-on experience with an LLM application involving retrieval, tool use or structured outputs, and the ability to explain its limitations. Comfort working with APIs, JSON, Git and code review, and taking a defined task from an agreed requirement to a maintainable implementation. Clear English communication and an evidence-based approach to debugging and evaluating system behaviour. Experience with Amazon Bedrock, Anthropic Claude, FastAPI, Docker, vector databases or self-hosted models is valuable. We value sound implementation decisions more than the number of frameworks on your CV. What you can expect You will contribute to real delivery work with direct access to experienced engineers and the founders. Expect clear ownership, thoughtful code review and a part-time scope planned around your studies. How to apply Use the Apply button on this page to submit your CV, current study details, location and weekly availability. Explain one AI application you built, your individual contribution, how you evaluated it and what you improved after testing. An anonymised description is welcome; please do not share confidential client material. We welcome applicants of all backgrounds. If the core responsibilities match your experience, we would like to hear from you. Stay connected Follow AgentixLake on LinkedIn for practical engineering insights, project updates and future opportunities. Following is optional and does not influence recruitment decisions. LinkedIn: https://www.linkedin.com/company/agentixlakegmbh/ Website: https://agentixlake.com/ Explore our production AI work with Claude: https://agentixlake.com/production-ai-with-claude