Know-how · ISLM

ISLM: Intelligent Scenario Lifecycle Management

Artificial intelligence delivers value in companies above all when its results do not remain isolated, but flow directly into business processes. This is exactly where Intelligent Scenario Lifecycle Management, or ISLM for short, comes in.

  • Learn how ISLM creates, integrates, and manages AI and machine learning scenarios in SAP S/4HANA.
  • Learn how traditional machine learning methods and generative AI are integrated into existing SAP business processes.
Request a free consultation on e-invoicing!
ISLM: Intelligent Scenario Lifecycle Management
Embedded
Side-by-Side
SAP HANA Machine Learning
Related Articles

Learn how artificial intelligence automates business processes, supports decisions and unlocks new potential for efficiency and innovation.

SAP Business AI Platform

Intelligent SAP processes with AI

Learn more →

Table of Contents

What Is ISLM?

ISLM stands for Intelligent Scenario Lifecycle Management.

It is a standardized self-service framework designed to manage the lifecycle of intelligent scenarios within SAP. An intelligent scenario represents a specific AI or Machine Learning use case within a business application.

Such a scenario can define:

which business data is used

which model or AI service is connected

which input values are required

which results the model returns

how the model is trained

how the model is deployed

and how its output is used within the business process

ISLM therefore connects the business use case inside SAP with the underlying AI or Machine Learning technology.

The core concept is to manage different AI technologies through a common framework and make their results available within SAP S/4HANA business processes.

What Does ISLM Stand For?

The abbreviation ISLM stands for:

Intelligent Scenario Lifecycle Management

The name reflects the main purpose of the framework: managing the complete lifecycle of intelligent scenarios in a structured way. Key lifecycle activities include:

  1. Creating an intelligent scenario
  2. Configuring input and output data
  3. Training a Machine Learning model
  4. Deploying the model
  5. Activating a model version
  6. Running inference or prediction calls
  7. Monitoring and updating the models in use

An important distinction is that ISLM is not a Machine Learning model itself.

Instead, the framework manages intelligent scenarios and connects a specific SAP business use case with the AI technology selected for that scenario.

What Are the Objectives of SAP ISLM?

The main purpose of SAP ISLM is to integrate artificial intelligence into SAP business processes in a structured way.

Rather than running AI capabilities only in external applications, companies can make predictions, recommendations or generated content available directly within an SAP application. Intelligent scenarios can, for example, be used to:

predict delivery delays

forecast sales or consumption volumes

classify documents automatically

detect anomalies in financial data

analyze quality notifications automatically

summarize business texts

generate recommendations for action

extract information from unstructured content

integrate generative AI assistants into SAP processes

The value does not come solely from generating an accurate prediction or producing text with AI. What matters is whether the result supports a concrete business decision or becomes part of an actual process activity.

How Does ISLM Work?

Intelligent Scenario Lifecycle Management acts as a standardized framework between the business process and the AI or Machine Learning technology used to support it.

Different technologies can be connected depending on the intelligent scenario. Examples include:

SAP HANA Automated Predictive Library

SAP HANA Predictive Analysis Library

SAP AI Core

Generative AI Hub

SAP BTP AI services

external Machine Learning services

customer-specific models

The appropriate architecture depends on the individual business use case. For example, a traditional forecasting model based on structured SAP data has different technical requirements from an LLM that is used to summarize longer business documents.

What Is an Intelligent Scenario in SAP?

An intelligent scenario describes a clearly defined AI or Machine Learning use case. It connects business requirements with the technical implementation.

An intelligent scenario can determine:

which data is required

which provider is used

which model is selected

which inputs are passed to the model

which outputs are expected

what the training data looks like

how the model is deployed

and how the results are subsequently used within the process

This enables companies to manage both intelligent scenarios delivered by SAP and their own customer-specific scenarios.

ISLM and Generative AI in SAP

In addition to traditional Machine Learning use cases, ISLM can also support Generative AI in SAP.

Through integration with Generative AI Hub in SAP AI Core, SAP applications can access Large Language Models. Generative AI Hub provides a central and controlled way of accessing different generative AI models.

This can make functions such as the following available within SAP S/4HANA:

summarizing longer business documents

generating suggested responses

extracting relevant information

classifying text

producing explanatory content

performing semantic analysis of business information

generating contextual recommendations

ISLM itself is not the underlying language model. Instead, the framework establishes the connection between SAP S/4HANA, the relevant business scenario and the AI infrastructure.

The model execution itself can, for example, take place through Generative AI Hub in SAP AI Core.

Which SAP Fiori Apps Are Part of the ISLM Framework?

The ISLM framework includes two central SAP Fiori apps. Each app covers a different part of the intelligent scenario lifecycle.

“Intelligent Scenarios” App

The Intelligent Scenarios app is used to create, configure, validate and publish intelligent scenarios.

Companies can use the app to:

display existing intelligent scenarios

create customer-specific scenarios

configure Embedded scenarios

configure Side-by-Side scenarios

assign models and providers

define input and output fields

define training and application data

validate scenarios

publish scenarios

The app therefore primarily represents the development and configuration level of ISLM.

SAP Fiori App ID: F4469

“Intelligent Scenario Management” App

The Intelligent Scenario Management app is used to operate intelligent scenarios after they have been published.

Depending on the scenario type, it can be used to:

train Machine Learning models

review training runs

deploy model versions

activate models

deactivate active model versions

remove deployments that are no longer required

manage access points for inference calls

This app therefore represents the operational and lifecycle management level.

SAP Fiori App ID: F4470

Embedded or Side-by-Side: Which ISLM Architectures Are Available?

Depending on where the business application and the Machine Learning provider are operated, ISLM distinguishes between two main scenario types:

Embedded scenarios

Side-by-Side scenarios

The architectural choice influences factors such as the technologies involved, data flows, required integrations and model operation.

What Are Embedded ISLM Scenarios?

In an Embedded ISLM scenario, the business application and Machine Learning functionality operate within the same technical stack. A typical example is SAP S/4HANA combined with the Machine Learning libraries of the SAP HANA database.

Relevant technologies include:

SAP HANA Automated Predictive Library (APL)

SAP HANA Predictive Analysis Library (PAL)

One advantage of the Embedded approach is that business data does not necessarily have to be transferred to an external platform. Calculations can take place close to the operational data within SAP HANA.

Embedded scenarios can be suitable for:

classification

regression models

forecasting

trend analysis

anomaly detection

predictions based on structured SAP data

What Role Does SAP HANA Machine Learning Play?

SAP HANA Machine Learning can be used as part of Embedded intelligent scenarios. APL and PAL are particularly relevant in this context.

SAP HANA Automated Predictive Library – APL

The SAP HANA Automated Predictive Library (APL) supports a largely automated approach to developing predictive models.

APL can handle different parts of the Machine Learning process, including:

selecting suitable modeling approaches,

processing relevant features.

The library is particularly suitable for users who want to create standardized predictive models without having to develop every individual modeling step themselves.

SAP HANA Predictive Analysis Library – PAL

The SAP HANA Predictive Analysis Library (PAL) provides a broad collection of statistical and analytical algorithms.

Compared with a more automated approach, PAL offers additional technical configuration options and can therefore be suitable for more complex or individually configured Machine Learning scenarios.

Its use generally requires a deeper understanding of the algorithms and data models involved.

What Are Side-by-Side Scenarios?

In a Side-by-Side ISLM scenario, the business application and the AI or Machine Learning provider run in different technical stacks. For example, SAP S/4HANA can run in its own system while the AI model is executed on SAP Business Technology Platform.

Possible providers and components include:

SAP AI Core

Generative AI Hub

SAP BTP AI services

external Machine Learning services

customer-specific models

specialized cloud or hyperscaler services

Side-by-Side scenarios can be particularly suitable for:

Natural Language Processing

document analysis

image analysis

Deep Learning

sentiment analysis

generative text processing

LLM-based assistance functions

customer-specific Python models

Model deployment is especially relevant in this architecture because the model runs on external infrastructure and subsequently needs to be made available to SAP S/4HANA.

Embedded vs. Side-by-Side in ISLM

Embedded

Execution environment: SAP HANA or SAP S/4HANA stack

Typical technologies: SAP HANA APL and PAL

Data: frequently structured SAP business data

Typical models: classification, regression and forecasting

Data transfer: often no transfer outside the stack is required

Strength: use cases close to SAP HANA data

Architecture: comparatively compact

Side-by-Side

Execution environment: external AI or Machine Learning infrastructure

Typical technologies: SAP AI Core, Generative AI Hub or external services

Data: structured and unstructured data

Typical models: LLMs, Deep Learning, document analysis and image analysis

Data transfer: depends on the architecture and use case

Strength: greater flexibility regarding models and services

Architecture: additional platform and integration components are required

ISLM vs. SAP AI Core – What Is the Difference?

ISLM and SAP AI Core are not competing solutions.

They perform different functions within a shared AI architecture. SAP AI Core provides a scalable runtime environment for AI and Machine Learning models on SAP Business Technology Platform. While SAP AI Core handles the technical execution and orchestration of models, ISLM connects those models with the business process in SAP S/4HANA.

In simplified terms, the responsibilities can be separated as follows:

SAP S/4HANA: Business processes and business data

ISLM: Definition and lifecycle management of the intelligent scenario

SAP AI Core: Execution and operation of AI models

Generative AI Hub: Controlled access to generative AI models

SAP AI Launchpad: Central management and monitoring of AI scenarios on BTP

ABAP AI SDK: Integration of generative AI into ABAP applications

How Does the Lifecycle of an Intelligent Scenario Work?

An ISLM project usually progresses through several lifecycle phases.

1

1. Define the Business Use Case

The starting point should be a business problem rather than a technology. Companies first need to define which decision or business process should be supported by AI.

A suitable use case requires:

a specific business problem

available and appropriate data

a measurable outcome

clearly defined process value

2

2. Choose the Scenario Type

The next step is to determine whether an Embedded or Side-by-Side scenario is more appropriate. Relevant decision criteria include:

type and volume of data

SAP version in use

required algorithms

Generative AI requirements

data protection

information security

scalability

existing BTP and SAP AI Core services

3

3. Configure the Intelligent Scenario

Within the Intelligent Scenarios app, elements such as the following are configured:

provider

model

input parameters

output values

For traditional Machine Learning scenarios, training and application data are also defined.

4

4. Publish the Scenario

Once configuration has been completed, the intelligent scenario is validated and published.

Only after publication can it be managed operationally through Intelligent Scenario Management.

5

5. Train the Model

For trainable Machine Learning scenarios, the model is created using historical data.

Training should be based on data that is suitable from a business perspective, complete and representative of the intended use case.

6

6. Deploy the Model

For Side-by-Side scenarios, the trained model is deployed on the relevant infrastructure.

This creates a technically accessible endpoint that can be used for later inference calls.

7

7. Activate the Model

A suitable model version is selected and activated for productive use.

The business application can then retrieve predictions or other AI-generated results.

8

8. Integrate the Result into the Business Process

This is the point at which the intelligent scenario generates actual process value.

A result can, for example:

be displayed in an SAP Fiori app

support prioritization

be passed to a workflow

be processed by a business rule

be shown as a decision proposal

be used for automation

9

9. Monitor Quality and Update the Model

Machine Learning models can lose predictive quality over time.

Possible causes include changes to:

data

products

customer structures

business processes

Companies should therefore regularly review model quality, model usage and the resulting business value. Where necessary, models can be retrained or replaced.

What Are the Benefits of ISLM?

ISLM provides a structured framework for managing different types of AI scenarios within SAP.

Standardized Lifecycle Management

ISLM standardizes activities such as creating, training, deploying and activating intelligent scenarios.

Direct Integration with SAP S/4HANA

AI-generated results can be used directly within SAP S/4HANA applications and business processes.

Support for Different AI Technologies

Organizations can connect SAP HANA-based Machine Learning models as well as external AI services and generative language models.

Centralized Management

Both SAP-delivered and customer-specific intelligent scenarios can be managed through a shared interface.

Reusable Integration

Connections to AI platforms and providers that have already been established can be reused across different application scenarios.

Governance and Control

Models and intelligent scenarios can be managed in a controlled way throughout activities such as:

publishing

deployment

activation

deactivation

Extensibility

Organizations are not limited to intelligent scenarios delivered by SAP. They can also create AI use cases tailored to their own business requirements.

What Are the Prerequisites for SAP ISLM?

The exact prerequisites depend on:

the SAP S/4HANA version

the operating model

the selected scenario type

Companies should typically review the following areas:

ISLM functionality available in the relevant SAP release

required SAP Fiori roles and authorizations

activated OData and UI services

SAP HANA revision

available Machine Learning libraries

technical connectivity to external AI providers

existing SAP BTP subaccount

configuration of SAP AI Core

access to Generative AI Hub

communication scenarios and communication systems

data protection requirements

information security requirements

authorization and governance concepts

Depending on the SAP product and release, connecting Generative AI Hub may require additional communication and configuration steps.

Does ISLM Require an Additional License?

According to the SAP product information contained in the source material, ISLM is generally provided as part of SAP S/4HANA.

Based on this information, the ISLM framework itself does not usually require a separate additional ISLM license. However, additional licenses or costs may arise for connected components. Examples can include:

SAP AI Core

SAP AI Launchpad

consumption of generative AI models

SAP BTP AI services

hyperscaler models

additional data services

additional integration services

Before implementation, companies should therefore review not only the technical availability of ISLM but also the licensing and consumption models of the AI services required for the planned scenario.

Which ISLM Use Cases Are Possible?

The available use cases depend on the business process, available data and selected AI technology.

ISLM in Finance

Possible use cases include:

detecting unusual postings,

forecasting payment developments

forecasting liquidity developments

classifying documents

assigning accounts to documents

summarizing financial discrepancies

generating explanatory texts

identifying information on invoices, for example through integration with OpenText Vendor Invoice Management

ISLM in Procurement

Potential scenarios include:

predicting delivery delays

assessing procurement risks

classifying purchase order texts

summarizing supplier communication

generating recommendations for action

ISLM in Sales

Potential use cases include:

forecasting closing probabilities

assessing quotation opportunities

predicting order delays

summarizing customer interactions

generating response texts

generating quotation texts

Manufacturing and Quality

Intelligent scenarios can, for example, be used to:

predict quality issues

classify defect patterns

suggest appropriate error codes

forecast material consumption

detect unusual production data

Service

Potential service scenarios include:

classifying incoming service notifications

summarizing long ticket histories

suggesting suitable resolution steps

identifying similar historical cases

prioritizing critical service processes

When Is ISLM the Right Solution?

ISLM: Intelligent Scenario Lifecycle Management can be particularly relevant when a company wants to:

use AI results directly in SAP S/4HANA

establish standardized processes for different AI scenarios

combine SAP HANA Machine Learning with BTP-based AI

create customer-specific intelligent scenarios

manage SAP and customer-specific models centrally

integrate Generative AI into ABAP and SAP processes in a controlled way

However, ISLM is not automatically required for every AI use case.

If an AI model operates exclusively outside SAP S/4HANA and does not require direct integration into an SAP business process, an architecture based solely on SAP AI Core or SAP BTP may be sufficient.

ISLM Best Practices: What Should Companies Consider?

Several key principles can be derived from the lifecycle and prerequisites described above.

Start with the Business Problem

The choice of AI technology should not be the first decision.

Companies should initially define which process needs to be improved and which specific problem they want to solve.

Review the Data Foundation Early

A suitable intelligent scenario requires data that is both available and appropriate from a business perspective.

Complete and representative training data is particularly important for trainable Machine Learning scenarios.

Distinguish Between Embedded and Side-by-Side

Not every use case requires SAP AI Core or an externally operated model.

Predictions that are closely connected to SAP HANA can be implemented as Embedded scenarios, while Generative AI and specialized models may require a Side-by-Side architecture.

Make Business Value Measurable

Every use case should have a measurable outcome.

Business validation can, for example, assess:

result quality

time savings

process improvements

Define Governance Before Production

Authorizations, monitoring, governance, support and controlled model activation should be addressed before the intelligent scenario is moved into productive operation.

Monitor Model Quality over Time

The lifecycle does not end when a model goes live.

Changes in data or business processes can also affect Machine Learning model quality. Models should therefore be evaluated regularly and updated when necessary.

Step by Step: How Do You Get Started with ISLM?

A structured implementation model can help organizations introduce ISLM in a controlled way.

1

Phase 1: Select the Use Case

Choose a clearly defined use case with:

measurable value

a suitable data foundation

a direct relationship to a business process

2

Phase 2: Define the Architecture

Decide whether the use case should be implemented through:

Embedded Machine Learning

a Side-by-Side model

a Generative AI scenario

3

Phase 3: Run a Technical Proof of Concept

Set up the required system connections and implement a limited prototype.

4

Phase 4: Validate the Business Results

Future users should evaluate the output of the intelligent scenario.

Relevant criteria can include:

result quality

time savings

process improvements

5

Phase 5: Move into Production

Before productive operation, organizations should establish areas such as:

authorizations

monitoring

governance

support

controlled model activation

6

Phase 6: Scale the Approach

After successful implementation, the technical and organizational approach can be transferred to additional business processes.

Conclusion: ISLM Connects AI Technology with SAP Business Processes

Intelligent Scenario Lifecycle Management provides a standardized framework for integrating artificial intelligence into SAP S/4HANA business processes in a controlled manner.

Companies can use ISLM to convert different AI technologies into structured intelligent scenarios. These scenarios can include traditional Machine Learning approaches based on SAP HANA as well as Side-by-Side architectures using SAP AI Core, Generative AI Hub and other SAP BTP AI services.

ISLM itself is neither a Machine Learning model nor a Large Language Model. Its role is to connect the business process in SAP with the relevant AI infrastructure. The main value therefore does not come from the use of artificial intelligence alone.

ISLM links the entire lifecycle of an intelligent scenario with the business process: from configuration and training to deployment, activation and productive use of the AI result.

For companies, this also means that the starting point should not be:

“Where can we use AI?”

A better first step is to identify a clearly defined business problem supported by an appropriate data foundation and measurable process value. Only then does it make sense to decide whether Embedded Machine Learning, a Side-by-Side model or a Generative AI scenario provides the most suitable architecture.

Consulting and Implementation of ISLM Scenarios

Fink IT-Solutions supports SAP customers in evaluating, designing and implementing intelligent scenarios.

Related Articles

Learn how artificial intelligence automates business processes, supports decisions and unlocks new potential for efficiency and innovation.

SAP Business AI Platform

Intelligent SAP processes with AI

Learn more →

Frequently Asked Questions About ISLM

What does ISLM stand for?

ISLM stands for Intelligent Scenario Lifecycle Management. It is an SAP framework for creating, integrating and managing AI and Machine Learning scenarios.

What is SAP ISLM?

SAP ISLM is a standardized lifecycle and integration framework. It connects intelligent scenarios with SAP business applications and supports the management of AI and Machine Learning scenarios within SAP S/4HANA.

Is ISLM part of SAP S/4HANA?

ISLM is provided as part of SAP S/4HANA or the underlying ABAP platform. The functionality available depends on the relevant release and operating model.

Is ISLM a Machine Learning model?

No. ISLM is a lifecycle and integration framework. The models themselves can be provided through technologies such as SAP HANA Machine Learning, SAP AI Core, Generative AI Hub or other AI services.

Can ISLM be used with Generative AI?

Yes. ISLM can be integrated with Generative AI Hub in SAP AI Core, allowing LLM-based use cases to be connected with SAP business processes.

What role does Generative AI Hub play in ISLM?

Generative AI Hub provides controlled access to generative AI models. ISLM establishes the connection between the business scenario in SAP and the corresponding AI infrastructure.

What is the difference between ISLM and SAP AI Core?

ISLM manages intelligent scenarios from the perspective of the SAP S/4HANA business process and connects them with business applications. SAP AI Core, by contrast, provides the technical infrastructure for executing and orchestrating AI models.

Can companies create their own intelligent scenarios?

Yes. In addition to intelligent scenarios delivered by SAP, companies can create and publish customer-specific intelligent scenarios.

What is the difference between APL and PAL?

APL supports more automated Machine Learning approaches and takes over different steps of model development. PAL provides a wider range of analytical algorithms and additional technical configuration options.

What is an Embedded ISLM scenario?

In an Embedded scenario, the business application and Machine Learning functionality run within the same technical stack. One example is SAP S/4HANA combined with SAP HANA APL or PAL.

When is a Side-by-Side scenario required?

A Side-by-Side scenario is used when the model or AI service operates outside the SAP S/4HANA stack, for example on SAP AI Core.

Can LLM results be displayed directly in SAP Fiori?

Yes. Through appropriate application development, generated content, summaries or recommendations can be displayed and further processed within an SAP Fiori application.

Which companies is ISLM suitable for?

ISLM is particularly relevant for companies that want to use AI results directly within SAP S/4HANA, manage different AI technologies through standardized processes, or integrate Generative AI into ABAP and SAP processes in a controlled way.

Does ISLM require an additional license?

According to the underlying information, ISLM is generally provided as part of SAP S/4HANA and usually does not require a separate ISLM license. Additional costs may, however, apply to connected components and services such as SAP AI Core, SAP AI Launchpad, generative AI models or other BTP services.

How much does an ISLM project cost?

The provided content does not specify concrete project costs. It states that additional licensing or usage-based costs may arise depending on the architecture and the external components or AI services involved. A specific cost estimate cannot therefore be derived from the source material.

How long does an ISLM implementation take?

The source material does not define a generally applicable project duration. It recommends a phased approach covering use-case selection, architecture design, Proof of Concept, business validation, production rollout and subsequent scaling. A specific timeline would require additional project information.

When is ISLM not required?

According to the provided content, ISLM may not be necessary if an AI model is used exclusively outside SAP S/4HANA and does not require direct integration into an SAP business process. In such cases, an architecture based only on SAP AI Core or SAP BTP may be sufficient.