SAP Datasphere as the Grounding Layer for Joule — Data Products, Access Controls and What Joule Reads Today

What you will learn

• Draw the four grounding surfaces of the SAP Business AI Platform (Business Data Tools, SAP Knowledge Graph, custom tools and JustAsk, document grounding) and state which ones Datasphere feeds and who owns each
• Choose between an SAP-managed data product and a replication flow you own for a given source, and design the flow's load type, privileges and monitoring for agent freshness
• State what Joule reads from Datasphere on 5 November 2026 (Joule in Datasphere, Analytical Insights, agents on SAP-managed data products, custom tools) and what is announced for later
• Run a data access control review that holds for an agent: identity in the permissions entity, enforcement by sharing into a consumption space, read logging, no copies outside Datasphere
• Place the four 'knowledge graph' products in their dated chain and promise only what ships this quarter
• Make one analytic model grounding-ready and prove the boundary with a test matrix

Course overview

SAP Datasphere is the modelling component of SAP Business Data Cloud, and it is where most of the governed data that an SAP agent or Joule answer rests on is shaped: data products installed into spaces, replication flows you own, analytic models with business semantics, data access controls that decide who sees which rows, and a catalog whose descriptions are what a language model reads. "Grounding Joule on Datasphere" is therefore not a toggle. It is a set of design decisions about which path data takes into the governed layer, which identity the reading component carries, which rows that identity may see, and how a business user's question finds the right model. The decision this course forces first is the identity one: an agent or tool that reads Datasphere with one technical user answers every asker with that user's scope, and no amount of prompt engineering changes that. (SAP Learning — Developing Detailed Knowledge of SAP Business Data Cloud)

Prerequisites

• SAP Datasphere foundations: spaces, Data Builder, analytic models
• “SAP Business Data Cloud End to End — Components, Personas and First Tasks” and SAP Business Data Cloud and data products
• “Joule — SAP's AI Engagement Layer: Which Joule, How a Request Runs, How to Roll It Out” helps to place the Agent Gateway and Analytical Insights

Outcomes

• Produce a grounding map for a client with the four surfaces, the Datasphere objects behind each and their owners
• Produce a source-by-source path decision (data product or replication flow) with load type, privileges and freshness owner
• Produce a data access control review record for one agent tool: identity, permissions entity entry, consumption space, audit retention
• Deliver the purchase-order lab with the four expected totals and the finding on the technical user
• Deliver a status ledger with the Datasphere AI features, the data product roadmap and the knowledge graph chain, dated