Get productive with Joule Work

Before you start working with Joule, it helps to understand the business context you’ll be using throughout this tutorial.

The scenario follows a Demand and Supply Planner at a discrete manufacturing company who needs to quickly review production actuals against the IBP forecast for the upcoming quarter, identify potential supply gaps, and decide where to act — all without switching between SAP S/4HANA, SAP IBP, and a series of manual reports.

This is exactly the kind of busywork BAIP and Joule Work are designed to eliminate. Rather than navigating reports and dashboards yourself, you’ll ask Joule directly in natural language and let it do the heavy lifting.

Throughout the steps that follow, keep this scenario in mind. Each interaction you have with Joule maps to a real task this person would otherwise spend significant time on.

The data used in this tutorial is sample data provided as part of the BAIP trial environment. It reflects a realistic but fictional business.

Now that you understand the scenario, put Joule to work.

👉 Open Joule Work Desktop and start a new conversation.

👉 Ask Joule to surface the actuals and forecast data for the upcoming production cycle. You can phrase this naturally: “Show me production actuals vs IBP forecast for Q3”.

Joule will immediately retrieve the data (actuals from SAP S/4HANA and the forecast from SAP IBP) and present it in a structured, readable format directly in the conversation. The forecast is predicted by leveraging SAP RPT 1.5, SAP’s tabular AI model, which uses your historical production data to produce highly accurate demand predictions.

👉 Take a moment to explore the result. Notice that Joule organizes the data into a visual card with the key metrics highlighted.

This step demonstrates a core capability of BAIP: Joule can access live business data through its connected data foundation and present it in context, without you navigating to a new system or opening a report.

A single conversation is useful, but Joule Work becomes even more powerful when you use Spaces — dynamically generated dashboards that automatically assemble relevant insights and data around a specific business goal or task in real time.

👉 In the conversation from the previous step, locate the actuals vs forecast card Joule generated.

👉 Save it to a new Space. Give the Space a meaningful name, such as “Q3 Production Review”.

👉 Open the Space. You’ll see the card pinned there, ready to reference at any time.

👉 Back in the conversation, ask Joule a follow-up question - for example, which region or product line is showing the largest variance — and add that card to the same Space.

Spaces are the connective tissue between conversations and ongoing work. They let you build up a curated view of a topic over time, across multiple interactions with Joule and with agents.

Spaces persist between sessions. When you return to Joule Work tomorrow, your Space and its cards will be exactly as you left them.

Joule Work goes beyond the traditional conversational interface. It serves as an orchestration layer for AI agents built on BAIP. In this step, you’ll invoke a specialized agent from within your Space.

👉 Open the “Q3 Production Review” Space you created in the previous step.

👉 With the Space active, start a new conversation and ask Joule to call the planning agent to analyze the forecast variance: “@planning-agent Analyze the Q3 variance data and suggest where we should rebalance production.”

Joule routes the request to the manufacturing agent, which processes the data and returns a prioritized set of production adjustment recommendations with projected impact.

👉 Add the agent’s output to your Space as a new card.

This step illustrates how BAIP connects the conversation layer (Joule Work) to the agentic execution layer underneath. The agent runs in the background; you stay focused on the outcome.

Joule allows you to continue, refine, and build on your conversations while maintaining context throughout.

👉 Return to the conversation from the earlier steps.

👉 Ask a follow-up question that builds on what’s already been discussed: “What are the current live production rates on Line 4?”

This time, Joule surfaces a data gap: the live production rate data isn’t available in the connected sources. Rather than estimating or fabricating an answer, Joule tells you exactly what’s missing and why it can’t respond fully.

This shows that Joule is grounded in real, connected data. When something is missing, it’s transparent about it and signals when it needs additional context.

Joule Work functions as a persistent, context-aware workspace rather than a stateless chatbot. Every interaction builds on the last, and the Space anchors everything to the business problem you’re solving.

Well done, you now have an overview of BAIP and Joule Work. For further info, see:

• Learn more about Joule Agents and Joule Assistants
• An overview of Joule Work
• Information on SAP-RPT