• Play a key role in building the internal platform which supports training and deploying AI models across the entire organization. • Take ownership of defining the platform to enable AI model fine-tuning and batch inference by building the SDKs and supporting the infra. • Manage cross-cutting stakeholder relationships, prioritizing customer needs first. • Navigate ambiguous, hairy, and technical problem spaces. • Jump into domains, languages, and problem areas that might be new and unfamiliar.

We are seeking a forward-thinking Full-Stack Developer to bridge the gap between robust SAP S/4HANA stability and the next generation of AI-driven enterprise applications. You will be responsible for building high-performance, Clean Core compliant extensions using SAP Business Technology Platform (BTP) and crafting intuitive front-end experiences with UI5, Angular. In this role, you won't just write code; you will actively integrate SAP AI Core and generative AI tools to accelerate delivery, automate complex business processes, and bring intelligent insights to the heart of the enterprise. What You’ll Do - Develop scalable, AI-driven applications on SAP BTP, leveraging SAP AI Core and AI Foundation services. You won't just build apps; you’ll architect systems that integrate Generative AI into core business processes (like Lead-to-Cash or Source-to-Pay). - Maximize productivity by using AI coding assistants (GitHub Copilot, Cursor) and SAP-specific LLMs to write, refactor, and document CAP (Cloud Application Programming Model) or RAP code. You are expected to stay at the forefront of prompt engineering for automated code generation. - Design and manage data pipelines that bridge the gap between structured SAP data and unstructured insights. This includes integrating RESTful APIs and managing high-performance databases, specifically utilizing SAP HANA Cloud’s vector engine for RAG (Retrieval-Augmented Generation) architectures. - Maintain a rigorous Git workflow to ensure stability in complex SAP landscapes. You will follow Gitflow strategies to manage clean commits and painless merges, ensuring that AI model configurations and application logic are versioned with enterprise-level precision. - Develop modular, reusable components in Angular (or SAP UI5) that integrate seamlessly with AI features. Your goal is to create "Agentic" user interfaces—UIs that don't just display data but provide AI-powered recommendations and natural language interactions. - Use Docker to containerize AI models and microservices, ensuring consistent execution environments. You will deploy and manage these workloads via SAP AI Core, ensuring your models are production-ready, resource-efficient, and compliant with SAP’s "Clean Core" strategy - Collaborate: Participate in code reviews (giving and receiving) and agile ceremonies. - Continuous Learning: Stay ahead of the curve by testing and adopting new AI dev-tools that improve the team’s workflow. What You Bring to the Table The Core Stack: - CAP (Cloud Application Programming Model): Strong foundation in Node.js or Java. Experience with the CAP framework is a must for building enterprise-grade services that integrate seamlessly with SAP HANA. - SAPUI5 / Fiori: Proficiency in JavaScript/TypeScript and modern SAPUI5. You understand the Fiori Design Guidelines and can build responsive, enterprise-ready frontends that users actually enjoy. - Git & CI/CD: You know the difference between a rebase and a merge, and you’re comfortable managing code versioning within SAP Continuous Integration and Delivery or Azure DevOps. - BTP & Containers: You can navigate the SAP BTP Cockpit, deploy MTAs (Multi-Target Applications), and understand why "Cloud Foundry" and "Kyma" (Kubernetes) are the backbone of modern SAP extensibility. The AI & Tooling Edge: - AI Orchestration: Experience using AI tools for boilerplate generation, unit test creation, and complex debugging. - Prompt Engineering: The ability to craft precise instructions for LLMs to get high-quality, secure code outputs. - Critical Evaluation: You don't just "Copy-Paste" AI suggestions; you have the technical depth to audit AI code for security vulnerabilities and performance bottlenecks. HITL (Human in the Loop) The "Nice-to-Haves": - Experience integrating OpenAI/Anthropic/Gemini APIs into web applications. - Familiarity with vector databases (like pgvector or Pinecone) or LangChain. - Familiarity with Cloud platforms (AWS, Azure, or GCP). - A "testing" mindset—experience with JUnit, or Jasmine/Karma or equivalent Who You Are - Pragmatic: You know when to write code from scratch and when to let an AI assistant do the heavy lifting. - Curious: You don't wait for a training manual; you're already experimenting with the latest LLM models and plugins. Quality-Focused: You understand that as AI generates more code, the human's role as a "Code Reviewer" becomes even more vital. What we offer - Flexible working hours and remote work - Innovative, interdisciplinary team environment - Employee benefits such as bonuses, cafeteria, Körber sports club, health fund, pension fund We are Körber – an international technology group with about 12,000 employees, more than 100 locations worldwide and a common goal: We turn entrepreneurial thinking into customer success and shape the technological change. In the Business Areas Digital, Pharma, Supply Chain and Technologies, we offer products, solutions and services that inspire. We act fast to customer needs, we execute ideas seamlessly, and with our innovations we create added value for our customers. In doing so, we are increasingly building on ecosystems that solve the challenges of today and tomorrow. Körber AG is the holding company of the Körber Group.

• Supporting an AI development team in SecTechOps • Building and testing agents • Developing use cases for further investment and adoption

• Advanced Prompt Engineering: Designing complex, dynamic prompt templates with conditional logic and efficiently reusing information and context within prompts to maximize generation quality and reasoning. • Structured Outputs & Schemas: Implementing various response schemes (JSON mode, function calling, Zod/JSON schemas) to ensure AI outputs are predictable and ready for seamless integration into application logic. • Prompt Engineering & Evaluations: Building robust evaluation pipelines and using Langfuse to collect feedback and score the quality of responses in real time. • Tracing & Debugging: Performing deep debugging of complex LLM chains using Langfuse traces to identify bottlenecks and optimize for cost, latency, and context window usage. • AI A/B Testing: Running systematic experiments across different models via OpenRouter (e.g., comparing Claude 3.5 Sonnet vs. GPT-4o) and analyzing results based on quantitative metrics. • Data-Driven Decisions: Making deployment decisions for new prompts or models strictly based on quantitative benchmarks and trace data, rather than intuition. • Output Scoring & Analysis: Developing scoring systems to analyze the “Problem → Solution” chain and identify root causes of hallucinations or logic errors using Langfuse analytics. • Model Performance & Fine-Tuning: Regularly re-evaluating model performance as new architectures emerge and performing fine-tuning when necessary to meet specific domain requirements.