Authored by Sarvesh Uttarwar, Innovation & Emerging Technologies, Körber Stellium
Most enterprise work is not difficult. It is scattered. A finance analyst knows the number exists somewhere in the system but not which report holds it. A new joiner needs to book leave and spends twenty minutes looking for the screen. A buyer who wants to know whether a supplier problem will affect next month’s production has to open four applications and assemble the answer by hand. None of this appears as a line item in any budget, and all of it adds up.
This is the problem SAP Joule is aimed at. Joule is SAP’s AI assistant, built into SAP applications rather than sitting alongside them, and it lets people ask for what they need in plain language instead of knowing where to look. It changes how people reach their systems rather than who does the thinking, and the practical effect is that the distance between a question and a usable answer gets considerably shorter.
What follows covers what Joule does, which assistants are available today and which are on the way, where it delivers value in the working day, what customers have reported, and what to put in place to get there.
- What SAP Joule actually is and how it works
- How Joule fits into your SAP enterprise landscape
- Key Joule Assistants and Agents: Capability and availability
- Use case: Managing supplier risk in manufacturing
- What to confirm before your SAP Joule rollout
- Where to begin with SAP Joule
What SAP Joule actually is and how it works
Joule is a conversational assistant that sits inside SAP applications and works from the user’s own role and permissions. SAP describes it as an assistant that knows your role. In practice it does three things: it answers questions using business data from SAP and connected third party sources, keeping the thread of a conversation so that follow up questions still make sense; it helps draft content, such as a job description; and it takes people into the right application with the relevant context already loaded, rather than leaving them to navigate there themselves.
It is worth separating two levels of Joule AI capability, because they are often discussed as though they were one thing.
Conversational assistance for everyday SAP tasks:
Finding information, summarizing it, explaining it and navigating to the right place. The user still performs the task.
Task execution with Joule assistants and agents:
An assistant or agent carries out steps inside a business process. SAP calls these Joule Assistants and Joule Agents, and the organization sets how much autonomy each one has. This is the newer of the two levels, and availability varies by product and agreement.
Both levels deliver value. Knowing which one a given capability belongs to is what makes a business case straightforward to build.
How SAP Joule has evolved to AI Agents
Two dates explain how Joule reached its current shape.
SAP announced Joule on 26 September 2023 as a natural language copilot to be embedded across its cloud portfolio. The announced rollout was staged: SAP SuccessFactors and the SAP Start site first, SAP S/4HANA Cloud public edition in early 2024, then SAP Customer Experience, SAP Ariba and SAP Business Technology Platform. At that stage Joule found information and explained it. SAP’s own example in the announcement has Joule identifying underperforming sales regions, connecting that to a supply chain issue, and then suggesting fixes for a person to review.
In October 2024 SAP introduced Joule collaborative AI agents, described as able to reason, plan and act, both individually and across more than one application. SAP’s illustration was a payment dispute, in which one agent takes the case, breaks it into steps and passes those steps to agents handling email, invoices and support cases. That was the point at which SAP moved from helping a person do the work to performing parts of it.
More recently, at SAP Connect in October 2026, SAP announced Joule Work, an interface for working across SAP and connected data grounded in SAP Knowledge Graph, a map of how business records such as orders, invoices and materials relate to one another. Alongside it came Joule Assistants that carry out operational tasks inside workflows using existing data, authorizations and audit trails. Joule Work is rolling out to all customers from October 2026, and the mobile app is already generally available.
How Joule fits into your SAP enterprise landscape
From a business perspective, three points are worth understanding.
First, Joule works through existing role based permissions, which simply means it inherits the access rights a user already holds and has none of its own. A user asking Joule a question sees only what their role already entitles them to see in the underlying application. This tends to shorten the first conversation with a security team.
Second, capabilities are presented as a single assistant rather than a separate tool for each product, so users do not have to work out which assistant belongs to which application.
Third, scope follows your landscape. SAP states that capabilities follow your SAP solution portfolio, that some features carry their own licensing, and that availability varies by agreement and region. Joule works across the applications and data sources connected to it, which is the purpose of grounding in SAP Business Data Cloud and SAP Knowledge Graph. Grounding means connecting an assistant to your actual business records, so that an answer comes from your data rather than from general knowledge.
Where Joule can reduce manual work
The clearest value is in work that is frequent, rule bound and spread across systems. Four examples, each with the problem, the way Joule addresses it, and the potential benefit.
Human resources and employee self-service:
Employees interrupt HR for tasks they could complete themselves, because the self service screens are used too rarely to become familiar. Joule lets an employee request or approve time off, or give and request feedback, through a conversation rather than a form. The benefit is fewer routine requests reaching HR and less time lost by employees. This was among the first Joule capabilities to ship, as part of the initial SAP SuccessFactors rollout.
Finance and transactional work:
Answering a question about a customer, an invoice or a balance usually means knowing which report holds it. Joule answers questions about business objects and takes the user into the relevant application with context loaded. SAP has also introduced assistants for specific finance work, including an International Trade Assistant for classification and an Accounts Receivable Assistant for collections. The benefit is faster access to information and less manual effort on repetitive classification and chasing.
Procurement and supplier decision support:
Sourcing decisions depend on supplier information held in several places. SAP has introduced a Joule Sourcing Assistant, and is embedding Moody’s supplier risk data into SAP Ariba so that risk signals appear within the sourcing process itself. The benefit is better informed sourcing with less preparatory research.
IT ,development and integration work:
Development and integration work carries a great deal of routine effort. Joule supports coding, testing, integration and application lifecycle activity. The benefit is more developer time spent on design and less on boilerplate.
Key Joule Assistants and Agents: Capability and availability
SAP separates assistants, which work with a user across a domain, from agents, which carry out specific tasks beneath them. At SAP Sapphire in May 2026, SAP described more than 50 domain assistants spanning finance, supply chain, procurement, human capital management and customer experience, drawing on a library of more than 200 specialized agents. The ones below are the most relevant to operations and finance teams, with availability as SAP has published it.
Selected SAP Joule assistants, use cases and avaliability
| Assistant | What it does | Availability |
| Supply Chain Planning Assistant | Senses, simulates and resolves planning exceptions in one workspace, including downstream impact | General availability planned Q4 2026 |
| Manufacturing Assistant | Helps teams respond to plant floor disruptions as they happen | General availability planned Q4 2026 |
| Logistics Assistant | Warehousing, outbound optimization, fulfilment and transportation dispatching | General availability planned Q4 2026 |
| Business Network Assistant | Sourcing, contracting and exception handling with suppliers and logistics partners | General availability planned Q4 2026 |
| Asset Operations Assistant | Starts, screens and simulates maintenance requests raised from equipment signals | General availability planned Q4 2026 |
| International Trade Assistant | Trade classification work in finance | In early deployment; SAP reports up to half the effort removed |
| Joule Sourcing Assistant | Sourcing support, with Moody’s supplier risk data surfaced in SAP Ariba | Novartis named as a pilot customer |
| Financial Closing Assistant | Automates journal entries, reconciliation and error resolution to shorten the financial close | Announced |
The table is also a planning tool. The supply chain assistants reach general availability in Q4 2026, which places them squarely in next year’s roadmap, while the trade and sourcing assistants are already running in customer deployments and pilots. Entitlement follows your commercial arrangement with SAP: RISE with SAP customers have three assistants activated within their first year, SAP GROW customers receive full portfolio access at onboarding, and SAP S/4HANA on premise and ECC customers receive selected AI scenarios where they commit to moving most of their landscape to SAP Cloud ERP.
SAP is also funding adoption. At Sapphire it announced a €100 million partner fund intended to accelerate deployment of SAP built AI assistants and agents. Eligibility sits with SAP partners, including those extending existing agents or building new ones on the SAP Business AI Platform using Joule Studio, so most customers reach it through a partner led engagement. Your account team or implementation partner is the route in.
Use case: Managing supplier risk in manufacturing
Consider a manufacturer with a few hundred direct suppliers. One of them runs into difficulty, and the business needs to answer a single question quickly: does this threaten next month’s production?
The information required is spread out. Supplier records and risk information sit with procurement. Open purchase orders sit in the ERP. The production plan sits in planning. In many organizations the first real signal of trouble is a late delivery, by which time the useful decision window has already narrowed.
What an experienced buyer does next is assemble the picture by hand: pull the open purchase orders for that supplier, identify which materials they supply, check which of those materials appear on the coming production plan, and establish whether an alternative source is qualified. It is careful work and it takes hours, and the answer begins to age as soon as it is finished.
Capabilities of the kind SAP has described could compress that sequence. Supplier risk information surfaced inside the sourcing application means the signal arrives where the decision is made rather than in a separate report. An assistant able to retrieve related records across procurement and operations reduces the manual assembly. Grounding in a knowledge graph is what allows connected records, such as orders and the materials behind them, to be brought together.
The business value is straightforward: faster access to the relevant information, clearer visibility of supplier related risk, less manual effort spent gathering it, and better informed procurement and production decisions.
The scenario is illustrative, assembled from capabilities SAP has published rather than from a single documented implementation. Its real use is as a test of fit: the more time your teams spend building that picture by hand today, the stronger the case for putting the information and the transaction behind one interface.
What early SAP Joule customers are reporting
SAP’s largest reference is its own business. SAP reports running Joule Work with 110,000 of its own employees, with productivity gains of around 20 percent across finance, HR and procurement. That is a deployment at full enterprise scale, and it shows what the technology does across three functions at once when the data underneath is in good order.
Maschinenfabrik Reinhausen offers the sharpest single example. SAP reports that work which previously took up to an hour of manual effort in SAP S/4HANA is now completed in seconds using natural language. That is what natural language access does to a well defined task.
On early deployments, SAP reports that its International Trade Assistant has reduced trade classification effort by up to half in early deployments. Novartis is piloting the Joule Sourcing Assistant, and Nestlé, PwC, Morgan Foods and CONA Services are each working with specific assistants. In supply chain, Doehler describes a Logistics Issue Resolution Agent that identifies root causes and recommends fixes.
Read together, these SAP reported results point to a clear pattern, and the pattern is more useful than any single number. The gains cluster in work that is high in volume, rule bound, and has a definite answer already sitting in a system: classification, lookup, reconciliation, chasing. That is a reliable guide to where to look first in your own operation.
The productivity case: Where SAP Joule productivity gains come from
Where Joule works, the gains come from a short list of effects: faster access to relevant business information, less time spent navigating applications, less manual effort on supported repetitive tasks, easier access for people unfamiliar with a particular screen or process, better visibility across related information, and more employee time available for work that genuinely needs human judgement.
How much of this an organization realizes depends on the use case, the quality of its data, the SAP applications and releases it runs, and how well employees adopt it. Every one of those is something a well scoped programme can shape, which is why the questions in the next section carry as much weight as the technology itself.
What to confirm before your SAP Joule rollout
Six questions, roughly in the order they tend to matter.
Existing SAP environment and release capability
.Which capabilities are supported depends on the SAP applications, releases and configurations you run. Confirm this against your actual landscape rather than a demonstration environment, and be specific about the release you are on.
Licensing and availability
Individual capabilities carry their own licensing, and availability varies by product, region, customer agreement and release. Confirm this with your account team as the business case is built, using the table above as the starting point.
Data quality and business data readiness
An assistant answers from your data, so the state of that data sets the quality of the answer. Clean supplier records, complete material master data and consistent classification between plants all raise it directly. Grounding works with what you have, which makes master data work done up front one of the highest return investments in the programme.
Security, permissions and AI governance.
Joule operates through existing role based permissions and audit trails, which settles the access question security teams raise first. The governance work that follows concerns actions: which steps an assistant takes without a person approving them, who owns the outcome, and how that is reviewed. SAP provides controls for setting autonomy levels per assistant and tooling for governing agents, and those are business decisions worth making early, with IT alongside.
User adoption and organizational readiness.
People act on an answer once they trust it. Budget for a verification period in which experienced staff check what the assistant tells them. That is how trust is built, and the teams that plan for it are the ones still using the tool a year later.
Designing a measurable Joule pilot
Choose one clearly defined business problem. Confirm the capability is available on your release and under your license. Establish a baseline for how long the work takes today. Give it an accountable owner. Then measure the result against that baseline.
Where to begin with SAP Joule
Joule is best understood as a change in how people reach their SAP systems rather than as a new system in its own right. For most organizations the first step is the same: identify the task that is highest in volume and lowest in judgement, confirm the relevant capability on your release, and run a measured pilot with a named owner and a baseline.
That approach pays back quickly. On work of the right shape the benefit shows up against the baseline within weeks, and it gives you a number of your own to build the next phase on. With more than 50 assistants now spanning finance, supply chain, procurement, HR and customer experience, and the supply chain set arriving in Q4 2026, the organizations that have already run that first pilot will be the ones ready to move at scale.
Your next step; From use case to measurable impact
The hard part is rarely the technology. It is choosing the right first use case and knowing what the capability will actually do once it meets your data.
Körber Stellium’s in-house Innovations team has built more than five assistants of its own across warehousing, manufacturing and transportation, each one shaped around a use case we identified in real operations rather than a product roadmap.
Talk to our experts to explore where SAP Joule fits in your landscape and scope a pilot worth measuring. 
