Join our team as a Custom Software Engineer, focusing on building AI-native, data-centric products within SAP BTP Datasphere. This role involves leveraging strong enterprise data warehousing and semantic modeling expertise, combined with agentic AI architectures like LLMs and tools. The goal is to move beyond traditional dashboards to create intelligent data experiences, including data agents, conversational analytics, and grounded insights.

SAP Datasphere serves as a robust data warehousing solution with extensive integration capabilities. We are looking for individuals who can design, develop, and enhance components across various systems and applications, utilizing modern frameworks and agile practices to deliver scalable, high-performing solutions tailored to specific business needs.

Design and implement governed data products on SAP BTP Datasphere, utilizing concepts like Spaces and shareable models. Build semantic models optimized for both analytics and AI consumption, ensuring clear definitions, measures, hierarchies, and lineage. Develop grounded AI experiences by connecting LLM applications to curated Datasphere models and enterprise sources, guaranteeing traceable responses. Engineer retrieval strategies that adhere to domain boundaries, freshness requirements, and access controls for reliable AI outputs.

Facilitate hybrid modernization and migration using SAP BW bridge patterns, supporting phased cloud transitions. Implement layered architecture designs (e.g., Bronze/Silver/Gold) for landing, cleansing, and publishing data, embedding quality controls throughout the process. Build data agents capable of planning, tool utilization, context retrieval, and generating cited answers, ensuring enterprise-grade safety and fallback mechanisms. Establish evaluation loops and telemetry for AI interactions to enhance reliability and efficiency.

A minimum of 2-5 years of experience is required for this role.

Essential skills include expertise in SAP BTP Datasphere for data modeling, spaces, sharing patterns, and enterprise semantic design. Strong fundamentals in data warehousing and the ability to translate business domains into governed analytical models are crucial. Hands-on experience with LLMs and RAG, encompassing retrieval, grounding, prompt and tool design, and evaluation is also necessary. Solid software engineering principles, including testability, a CI/CD mindset, and reliable integrations, are fundamental.

Beneficial skills include migration and modernization experience with BW bridge patterns, layered architecture implementation (Bronze/Silver/Gold), and familiarity with vector search/embedding pipelines for integrating external AI retrieval components. A 15-year full-time education is a mandatory qualification.