Enterprise Artificial Intelligence

An Introduction to SAP’s AI Strategy

By Dr. Feiyu Xu, Global Head of Artificial Intelligence, SAP and Maximilian Herrmann, Project Consultant Artificial Intelligence, SAP

Leveraging artificial intelligence (AI) to add business value is no longer a vision of the future for enterprises. Applying AI — and applying it in the right way — is necessary for businesses that want to predict and shape future outcomes, empower their employees to perform higher-value work, and automate and optimize business processes, decisions, and experiences today. Companies that have not invested in an AI strategy of their own are in danger of missing out on opportunities that their competitors are already taking advantage of, such as improving their products, innovating new products, increasing revenue, improving profit margins, and reducing risks.

Organizations across various industries use AI as a catalyst for business process agility, digital transformation, and accelerated innovation. If AI is used and applied correctly, it can offer transformational possibilities for SAP customers, businesses, and overall society. Consider, for example, an article published in late 2020 by Jana Wuerth and Ivona Crnoja from the SAP AI team, which demonstrates how AI business services fit into a comprehensive digital process automation strategy and how these processes can help SAP customers drive end-to-end process automation.

SAP AI strategy author headshot

SAP AI Strategy Author bio

While many organizations are still investing in creating their own AI models to support their unique needs, most organizations will experience the need for AI as functionality that gets embedded within a packaged application. However, adopting AI doesn’t come without challenges. Research conducted by O’Reilly Media in 2019, for example, examined the following question: what is the main bottleneck holding back further AI adoption? The results showed that most respondents (23%) think the bottleneck lies in a company culture that does not yet recognize the needs for AI. Other respondents identified lack of data or data quality issues (19%), lack of skilled people (18%), and difficulties in identifying appropriate business use cases (17%) as the main challenge holding adoption back.

These are just a few of the challenges that companies face while implementing AI. Nonetheless, enterprises need AI applications to ensure integrated processes across all lines of business (LOB), and to keep pace with a growing AI market. According to a study conducted by the International Data Corporation (IDC), worldwide revenues for the AI market, including software, hardware, and services, are forecast to grow 16.4% year over year in 2021 to $327.5 billion. By 2024, the market is expected to break the $500 billion mark with a five-year compound annual growth rate of 17.5%. According to the same study, worldwide revenues for the AI software market alone reached $248.36 billion in 2020 and are forecast to grow with a five-year compound annual growth rate of 17.3%. Furthermore, most major software vendors invest aggressively in integrating AI into their applications and consider AI to be the core architectural application infrastructure.

Considering this steady growth and SAP’s investment in AI technology, this article aims to help enterprises drive their AI strategy, beginning with a deep dive into the key aspects of SAP’s own AI strategy. This explanation is followed by two examples of how companies can leverage AI that SAP has already embed- ded into some of its enterprise applications and other functionality that SAP has to offer.

Understanding SAP’s AI Strategy

When implementing AI, SAP customers often face high effort and failure rates due to complex custom development projects. Successfully implementing AI is more complicated than simply implementing a single traditional application or use case — and yet, customer needs are precise: They want AI to be part of their portfolio, and SAP’s vision to offer a holistic solution portfolio for every industry and across all end-to-end processes covering lead-to-cash, design-to-operate, source-to-pay, and recruit-to-retire — and injecting AI into all of these applications and business processes — is a step forward to help companies become an intelligent enterprise.

SAP offers products and services for multiple lines of business (LOB) and various industries to support the different business processes that keep enterprises running. SAP has been working on enterprise AI solutions for many years and has now infused AI directly into its business processes, enabling customers to use AI functionality natively in SAP solutions as part of their standard applications.

How does SAP embed AI into all these processes and applications, and how do we develop ready-to-use AI functionalities internally?

The internal AI Factory approach — an efficient and holistic development approach created at SAP (Figure 1) — covers the entire development process from ideation and innovation to productization, go-to-market (GTM), and continuous improvement (Figure 2). Our goal: make the transformation to an intelligent company as easy as possible for our customers. Every relevant stakeholder is involved in this process.

With this approach, SAP supports various LOBs and industries to help them understand, contextualize, and apply AI capabilities at scale. Our internal AI Factory approach consists of assets, tools, and a holistic collaboration model. The internal technology component that provides our AI infrastructure is called the AI foundation. Additionally, SAP sees AI functions pro- viding centralized development and management of data corpora, AI general, and business functions (i.e., application program interfaces, pipelines, and models) and common key performance indicators (KPI).

The goal of SAP’s AI technology is to bring value to SAP customers, such as cost savings, efficiency improvement, and better human-machine interaction. However, the past has shown that realizing business value from AI does not happen automatically — it is reliant on an interplay of many complex factors. Therefore, SAP’s AI strategy is an integral part of SAP’s product strategy that ensures technologies, applications, adaptions, GTM, legal framework, collaboration with partners, technology providers, data providers, and the academic community come together in one joint collaboration process to embed AI into SAP’s portfolio.

SAP AI Strategy Figure 1
Figure 1: SAP’s holistic approach to putting AI into production: AI Factory, AI-supporting functions, and AI ecosystem
SAP AI Strategy Figure 2
Figure 2: SAP’s internal AI Factory approach covers the entire development process from ideation and innovation to productization, GTM, and continuous improvement

In summary:

■ SAP is embedding AI into its applications and injects AI into business processes in a scalable way.

■ The internal AI Factory approach at SAP is a scalable AI application development approach pro- viding assets, tools, and a collaboration model. This approach enables SAP to put AI into production covering all steps of efficient application development and all relevant stakeholders.

■ Also, SAP exposes AI capabilities via the SAP Business Technology Platform, so customers can create their own extensions.

■ SAP does not compete in general-purpose AI services and platforms. Our AI services and platforms are close to SAP solutions and customer business processes.

■ SAP offers a true evolution of enterprise AI.

What Does Embedded AI Mean?

The term embedded AI refers to the use of machine and deep learning capabilities that exist inside (or are embedded within) a software product. AI functionality is embedded in many SAP solutions as part of the standard application in the form of pre-trained models, like ExpenseIt in SAP Concur, or models that are trained using customer data, like SAP Cash Application. SAP also offers a broad and deep portfolio of enterprise applications that enable customers and business users to run their business processes with embedded AI and configure and adopt AI as part of their standard workflow.

SAP Conversational AI and SAP AI Business Services are part of the AI Factory and are well-integrated into the intelligent suite. SAP Conversational AI is SAP’s low-code no-code chatbot building platform to optimize customer service in conversation-driven interactions with end-customers. It offers a powerful language technology, an end-to-end platform to train, build, and monitor chatbots.

SAP AI Business Services provides strategic AI capabilities that automate and optimize business processes to enrich the customer experience across the intelligent suite. They are provided as reusable services and applications that are optimized for SAP solutions on the SAP Business Technology Platform Core Extension Suite. Using SAP’s Artificial Intelligence Business Services, customers benefit from innovations developed within SAP’s ecosystem. They can identify and elaborate new use cases, extract and prepare data for analysis, and integrate SAP’s business services into their business processes.

SAP Intelligent Robotic Process Automation and SAP Data Intelligence are also some of the building blocks integrated into the intelligent suite. Combining these intelligent technologies helps SAP to embed AI into its integrated business processes and allows customers to extend the intelligent applications suite with new intelligence or create new intelligent solutions.

You may have noticed that AI is already in the core of enterprise software (Figure 3). For example, machine learning, analytics, and data intelligence functionalities are currently part of SAP HANA and SAP S/4HANA products. SAP is now accelerating and broadening its AI efforts across all product portfolios. AI at SAP has a solid infrastructure for training and inferencing and lifecycle management, continuing with various core machine learning technologies such as AutoML, explainable AI, then business-driven natural language processing and yielding in applications for analytics, prediction, optimization, digital assistants via chatbot, and dialogue technologies.

Examples of AI-powered SAP applications include:

SAP Service Cloud: Helps each agent save up to 40 minutes per day by automating 70% of the ticket management process. Combines ticket classification, automated answering, and natural language processing.

Sales Order Processing: Saving approximately 18 million euros per year (350 FTE) for one customer by automating filling of 12 million fields per year where rule-based approaches failed. The processing uses SAP Data Attribute Recommendation service.

Document Information Extraction (SAP S/4HANA, CX, Concur): Can result in a decrease of time spent on invoice processing by approximately 35%. Can also process various other documents, e.g., automatically process payables and payment notes and ensure that invoices and payables match.

SAP AI Strategy Figure 3
Figure 3: SAP embeds AI in standard SAP applications and offers enterprise-specific AI solutions and services to extend and integrate business processes
SAP AI Strategy Figure 4
Figure 4: SME Document Processing — Invoice Scanning: Example of efficient collaboration and timely market introduction and adoption

SAP’s embedded AI can add value to various LOBs and industries. The following example drills down into how embedded AI can add value to businesses’ invoice processes specifically.

Using Embedded AI to Increase Efficiency Around Supplier Invoice Processing

Entering supplier invoices can be a time-consuming job, especially when massive data must be captured manually from a paper invoice or multiple PDF documents. With embedded AI, business users can take advantage of advanced capabilities — like scanning PDF documents using SAP’s intelligent document information extraction service — to process supplier invoices more efficiently and accurately. Figure 4 illustrates a positive example of how SAP applies the internal AI Factory approach to intelligent supplier invoice scanning for small businesses and medium enterprises (SME). Demonstrated in Figure 4 is a classic invoice scanning scenario in ByDesign for which SAP Document Information Extraction is being used. Customers upload invoices, and SAP’s service automatically extracts key information and pre-fills the relevant fields in ByDesign.

The SME and the Technology & Innovation (T&I) board’s AI team partnered together to build this application. From ideation over productization to delivery, it took only six months in total, and in September, it already had more than 135 customers. However, what made this scenario successful is that the SME has the overall ownership and focuses on the identification of the use cases, customers, market size, proof-of-concept, productization, delivery, and GTM. In the GTM phase, they leverage onboarding automation tools to ensure a timely adoption. The AI team supports reusable technologies, AI Business services, adaptation to SME needs, and lifecycle management.

Conclusion

SAP is committed to helping businesses inject AI into their business processes in a scalable way to help them realize business value. Learn more about AI and ML at SAP by visiting the SAP Community Page and read more about how AI solutions from SAP can help enterprises solve complex business challenges. You can also find out more about SAP AI Ethics and how SAP is supporting AI for the good. For more information about how SAP is engaging in active AI research, visit SAP AI Research.

The authors of this article would like to acknowledge Joelle Eline Weber, T&I AI Operations, and the support she provided to create this article.