Many SMEs continue to operate their ERP landscape on-premise - for example with SAP ECC or locally operated S/4HANA. These systems have grown over the years, are deeply integrated into existing processes and run stably. A complete transformation to a cloud ERP environment therefore often means
extensive migration projects
high investment costs
long project durations
For medium-sized companies in particular, such a transformation is often not a short-term project. At the same time, many new AI functions in the SAP ecosystem are primarily created in cloud-based solutions. This creates a challenge: companies want to use AI - without having to transform their entire ERP landscape immediately. The good news: AI in the SAP environment is also possible with existing on-prem systems.
Getting started with AI does not necessarily have to begin with a complete cloud migration. Companies can combine various approaches to integrate AI into their processes today.
Even if the central ERP system is still running on-premise, companies can already use AI functions in SAP cloud solutions. Examples include:
SAP SuccessFactors for HR processes
SAP Sales Cloud and Service Cloud in the customer experience environment
These solutions already offer integrated AI functions - for example for
intelligent recommendations
Automated analyses
This allows companies to use AI-supported functions without having to transform their entire ERP system.
Options for data-driven analyses and initial AI-like functions already existin S/4HANA On-Premise . These include, for example
Machine learning applications directly on the HANA database
Embedded analytics
Situation handling for the automatic detection of critical events
Intelligent Fiori apps
These functions already enable automated analyses and proactive decision support within the ERP system.
Companies that also use the SAP Business Technology Platform (BTP) can further expand their AI capabilities. Various services are available here, for example for
Document processing
machine learning
Process automation
Generative AI integration
BTP often acts as a bridge between existing ERP systems and new AI services.
In addition to the integrated options, many companies rely on individual AI applications that are specifically tailored to their business processes. One tried-and-tested approach is an AI use case ideation workshop, for example. This involves working with specialist departments and IT to analyze specific issues, such as
Where do a particularly large number of manual tasks arise?
Which decisions are based on large amounts of data?
Where could forecasts or automation create real added value?
The aim is to identify AI use cases with a clearly measurable business impact instead of just introducing new technologies. Depending on the system landscape, these use cases can then be implemented directly in SAP, via extensions or via individual AI services.
Find out more about AI use case workshops now
Generative AI offers great potential in areas with a lot of data, documents and recurring tasks. These primarily includeSupply Chain, Finance, Procurement and Sales.
The supply chain is becoming increasingly complex: suppliers, stock levels, transportation routes and customer requirements generate large amounts of data every day. Generative AI can help to make processes more intelligent, identify risks at an early stage and improve decisions. This is achieved through, among other things:
Intelligent demand forecasts based on the analysis of historical sales data, seasonal patterns and external factors
Automated warehouse and inventory optimization by monitoring stock levels in real time and detecting bottlenecks
Evaluation of supplier stability and reliability
Dynamic route planning
Prediction of delivery times and delays
Support in choosing cost-efficient transportation options
Generative AI makes the digital supply chain smarter, faster and more resilient. This makes the supply chain not only digital, but also intelligent and proactively controllable.
Finance departments work with large volumes of data every day - from invoices and bookings to reports. AI can make many processes much more efficient here. Typical use cases include
Automated invoice processing
Analysis of incoming invoices, data extraction and creation of booking proposals
Automatic financial analyses, asanomalies in financial data can be recognized and deviations can be explained automatically
Forecasting and planning
By analyzing historical data , AI can create forecasts for sales, costs or cash flow. This reduces manual work and at the same time improves the decision-making basis for finance teams.
Procurement also generates large amounts of data on suppliers, prices and contracts. AI supports purchasing departments in areas such as
Analyzing supplier risks
price forecasts
Automatic order proposals
contract analysis
This enables companies to react more quickly to market changes and identify risks at an early stage.
Generative AI is also creating new opportunities in sales. Examples include
Customer analysis:identifying patterns in purchasing behavior and cross-selling and upselling potential
Automatic offer drafts
AI-supported sales support: draft offers can be created based on customer requirements
Sales employees automatically receive relevant information on customers, products or sales opportunities. This reduces administrative tasks and creates more time for customer care.
A key success factor for AI projects is the database. In many companies today, data is still stored in different systems and silos. Without a clear data strategy, there is a risk that individual isolated AI use cases will emerge that are difficult to scale in the long term. This is why many organizations rely on modern data platforms and target architectures that bring together data from different systems, provide consistent data models and make AI applications scalable
A solid data architecture therefore forms the foundation for sustainable AI strategies.
Even companies with on-premise systems can prepare for future AI innovations today. An important approach here is the clean core principle. The aim is to keep the ERP system as close to the standard as possible and implement extensions via modern platforms or services. This facilitates later system updates, cloud transformations and the use of new AI functions in future SAP releases.
Clean Core thus becomes an important building block for a future-proof and AI-driven ERP architecture.
In addition to SAP-based AI scenarios, AI functions are also becoming increasingly important in the Microsoft world. Microsoft 365 Copilot opens up new possibilities for
automated document creation
intelligent meeting summaries
Analysis of company data
Support in daily work processes
In many companies, this creates combined AI scenarios in which SAP data, business processes and modern workplace tools are linked together.
For many SMEs , the most practical way to get started with AI is currently via specialized solutions. This approach enables companies to optimize individual processes in a targeted manner, continue to use existing SAP systems and gradually introduce artificial intelligence. At the same time, companies still have the option of further developing their own system landscape in the long term - for example as part of a later cloud transformation. AI therefore does not become a major transformation project, but rather a gradual further development of existing business processes.
Many SMEs are at a crucial point today: they have come fromstable on-premise worlds and recognize the opportunities offered by the cloud and AI, but the big, immediate leap often seems neither realistic nor necessary. This is precisely why radical upheavals are not needed, but rather well thought-out, step-by-step approaches. Solutions that respect existing ERP landscapes, but at the same time pave the way towards AI and modern architectures. The key is to meet SMEs where they are today: with a clear strategy and a realistic transformation path.
AI in the SAP midmarket does not have to be a major project. On the contrary: the greatest added value often comes from focused, pragmatic use cases that quickly make an impact and can be seamlessly integrated into existing systems. If you want to be successful, focus on:
concrete business problems instead of technology hype
small, measurable steps with a clear ROI
Hybrid scenarios instead of "all-or-nothing" transformations
Solid database as a foundation for AI
and above all: the involvement of people in the company
The result is not a disruptive break, but a sustainable development. For SMEs, the path to the AI-supported SAP world is therefore not a question of "if", but of "how" - and this "how" does not begin with a major project, but with the first sensible step.