Analytics Legends The knowledge platform for SAP Analytics
Academy module

Natural-Language Query over SAP Data — Text-to-SQL Patterns and Limits

Natural-Language Query over SAP Data — Text-to-SQL Patterns and Limits — Analytics Legends section illustration for the SAP Analytics knowledge base (concepts, studies, Academy)

As of 2026-09-25

Edited by , sole editor and publication director — 27 years in the SAP ecosystem. How these pages are produced, and where automation is used.

Compares the three real natural-language-to-query systems a consultant meets in an SAP landscape — SAP Analytics Cloud's Just Ask (business user, 2024 Q1 QRC, SAC + Datasphere models, replaces search-to-insight), Joule's SQL generation in SAP HANA Cloud's developer tooling (a coding assistant producing SQL a human must review, not a business-user query), and Databricks Genie inside SAP Business Data Cloud (Unity-Catalog-governed, with documented configuration layers: descriptions as 'mission statements', synonyms/vocabulary, value dictionaries, trusted functions). Uses Databricks' own published best-practice guidance to name the shared failure modes — ambiguous scope, join-cardinality double-counting, absence of a benchmark — and gives a four-rule cross-platform playbook: know the audience, benchmark before configuring, fix logic drift with trusted functions, and always treat write-capable generated SQL as a reviewed draft. Three exercises classify real questions, build a benchmark set, and diagnose a join failure. Builds on M346 and M347; feeds M349.

What you will learn

  • Distinguish SAP Analytics Cloud's Just Ask, Joule's SQL generation in HANA Cloud, and Databricks Genie by target audience and whether raw SQL is ever exposed
  • State Just Ask's documented scope and release (2024 Q1 QRC / 2024.02, SAC + Datasphere models) and that it replaces search-to-insight when enabled
  • Explain why Joule-generated SQL must always be treated as a reviewed draft, never auto-executed, against anything beyond a disposable schema
  • Map Databricks Genie's three configuration layers (descriptions, synonyms/vocabulary, value dictionaries) plus trusted functions to the equivalent SAP semantic-layer disciplines from M347
  • Build a benchmark question set with ground truth and explain why it must exist before any text-to-SQL configuration change is judged

Module overview

Who this is for. You have seen a demo where someone types a question and a chart appears, and a client now expects you to explain, precisely, what happened between the two. This module compares the real natural-language-to-query systems a consultant meets in an SAP landscape — SAP Analytics Cloud's Just Ask, Joule's SQL generation inside SAP HANA Cloud's developer tooling, and Databricks Genie inside SAP Business Data Cloud — and gives a shared vocabulary for their limits. It builds on M346 and M347.

Prerequisites

  • M347 (The Semantic Layer as LLM Context) or working knowledge of analytic models and business-layer design
  • M346 (AI-Ready Data) or understanding of catalog metadata, glossary terms and authorisation scoping
  • Familiarity with SAP Analytics Cloud and, ideally, exposure to a Databricks or SAP Business Data Cloud environment

Outcomes

  • Correctly route a client's natural-language query need to Just Ask, Joule SQL generation or Databricks Genie based on audience and risk.
  • Configure (or specify for a client's team) the metadata layers a text-to-SQL system actually reads, not the prompt wording.
  • Build and defend a benchmark question set as the basis for judging any configuration change.
  • State and enforce the human-review rule for any generated SQL capable of writing data.

Full module available to members. The full module adds: the decision framework · the end-to-end scenario walkthrough · the KPI scorecard · the anti-patterns · the code blocks · the knowledge check · the diagrams.

Open in the app →

Next in this track