Embeddings and vector search from Datasphere and HANA Cloud tables

Where the vector capability actually lives

A common misreading is to look for "Datasphere's vector store" as a distinct product. It is not one: the vector capability sits in SAP HANA Cloud, the database Datasphere is built on top of. SAP's own description of the HANA Cloud vector engine names three things it does: generate a vector embedding from text, run a nearest-neighbour search using cosine similarity expressed directly in SQL, and enrich the results by joining the vector-search output with ordinary relational business data. That third point is the one worth sitting with: a similarity search here does not hand back an isolated list of matches the way a standalone vector database would — it returns rows that can be joined, in the same query, to the measures, hierarchies and business context that already live in a Datasphere-modelled table.

The SQL you will write

HANA Cloud stores embeddings in a REAL_VECTOR(<dimensions>) column. VECTOR_EMBEDDING(<text>, 'DOCUMENT' | 'QUERY', '<model>') computes an embedding inside the database with an SAP-provided model, so no text leaves HANA Cloud. Similarity is COSINE_SIMILARITY(a, b) or L2DISTANCE(a, b), used in ORDER BY … LIMIT k; for large tables a CREATE HNSW VECTOR INDEX makes the search approximate and fast. In a Datasphere-modelled landscape, run this in an open SQL schema or a SQL view over the space's tables, and join the result back to the analytic model's dimensions. Check the vector engine guide for the models and dimensions available in your release.

Why it matters

• SAP's vector engine joins similarity-search results back to relational business data in the same SQL query — a structural difference from a standalone vector database that changes what "grounding on your own tables" can deliver.
• Embedding structured business data and document grounding are different pipelines with different tooling; scoping them as the same work under-delivers on both.
• Where an embedding lives — HANA Cloud versus a Databricks-side vector search — is a governance decision as much as a technical one, since it decides whether a second, ungoverned copy of the data gets created.

Key points

• The vector capability sits in HANA Cloud, the database beneath Datasphere — not in a separate Datasphere-specific vector product.
• SAP names three vector-engine capabilities: text-to-embedding generation, nearest-neighbour search via cosine similarity in SQL, and joining vector results back to relational business data.
• This page covers embedding structured table data (product descriptions, feedback, incident notes); document grounding (PDFs, SharePoint) is a separate pipeline.
• Keeping embeddings inside HANA Cloud preserves governance — the vector stays in the same access-controlled place as the row it was derived from.
• SAP pairs the vector engine with a knowledge-graph engine on the same database for hybrid retrieval: traversal, SQL aggregation and vector similarity without moving data.
• Choose HANA Cloud vs. Databricks Mosaic AI vector search based on where the consuming application runs, not on raw capability.
• Mixing outputs of two different embedding models in one index silently breaks similarity search — a common, avoidable design error.

Terms used on this page

HANA Cloud vector engine

SAP HANA Cloud's capability to store real vector data, generate text embeddings, and run cosine-similarity nearest-neighbour search in SQL, joinable to relational data.

Structured-data embedding

Turning a text column already stored in a governed business table into a vector alongside its row, as distinct from embedding documents.

Hybrid grounding

Combining vector similarity search with knowledge-graph traversal and SQL aggregation on the same database.

Embedding model mismatch

The error of mixing vectors from two different embedding models in one index, which silently breaks similarity comparisons.

Cosine similarity

The distance metric HANA Cloud's vector engine uses to rank nearest-neighbour matches directly in a SQL ORDER BY clause.

Refresh cadence (embeddings)

The schedule on which a stored embedding is regenerated to stay current with changes to its source text column — a one-time batch job goes stale silently without one.

Cortex Search (peer term)

Snowflake's managed hybrid (vector + keyword) search service over tables, positioned as automatically keeping its index synchronized with source-table changes.

Delta Sync Index (peer term)

The Databricks AI Search (the renamed Vector Search) index type that stays automatically synchronized with a source Delta Lake, streaming, or Iceberg table (Databricks documentation).

Sources

1. SAP — Run semantic search on data with the SAP HANA Cloud Vector Engine (text embedding, cosine-similarity nearest-neighbour in SQL, joins to relational data)
2. SAP Community — what's new in SAP HANA Cloud, September 2026
3. SAP Community — querying RDF graphs with SAP HANA Cloud Knowledge Graph Engine (graph retrieval next to vector search)
4. SAP — HANA Cloud vector engine and Knowledge Graph engine asset page
5. Databricks documentation — AI Search (the renamed Vector Search) and Delta Sync Index (fetched 2026-09-27)
6. SAP AI Core — generative AI hub and AI Units metering
7. Embeddings & Semantic Search — SAP Architecture Center
8. Retrieval Augmented Generation using generative-ai-hub-sdk and HANA vector search — SAP Developer Center tutorial
9. Using the SAP HANA Cloud Vector Engine with the SAP HANA Cloud Knowledge Graph Engine — SAP Help Portal
10. SAP HANA Cloud Vector Engine announcement — SAP Community (Nov 2023)
11. Embedding Business Context with the SAP HANA Cloud Vector Engine — SAP Community
12. Mastering SAP HANA Cloud Vector Engine for RAG-driven LLM Applications — SAP Community
13. cap-ai-vector-engine-sample (CAP, AI Core and HANA Cloud vector embeddings) — SAP samples on GitHub
14. Using Advanced AI Techniques with SAP's Generative AI Hub (document grounding course) — SAP Learning
15. Orchestration(V2) with Grounding Capabilities in SAP AI Core — SAP Developer Center tutorial
16. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks — arXiv (Lewis et al.)
17. SAP HANA Cloud, SAP HANA Database Vector Engine Guide — SAP Help Portal
18. SAP TechEd 2026 press release (27 Oct 2026) — Joule Studio availability