This course uses a financial services use case to build a complete AI context store and RAG pipeline on SAP HANA Cloud Vector Engine, designed and implemented from first principles using a native implementation, connecting directly to HANA without going through AI Launchpad or GenAI Hub. The course opens with a live demo using the chat client. We ask actual analyst questions with and without grounding and set the stage for the complete design and development of the context store.
We spend time understanding how SAP HANA Cloud is the foundation context store for enterprise AI including Joule, AI Core, custom agents, and MCP all ground their answers here. We cover vectors, what they are, how VECTOR_EMBEDDING() computes them inside HANA, the difference between DOCUMENT and QUERY embedding types, and how cosine similarity measures semantic distance.
We then analyze the document corpus — a credit risk policy, an earnings report, and a portfolio risk register. We design the context store schema and build the table with full lifecycle support including embeddings, document identity, versioning, section filters, governance flags, and audit metadata. We configure the HNSW vector index — build and search parameters, what each controls, and the exact syntax HANA requires.
