In recent years, we've seen a surge in interest and application of Artificial Intelligence (AI), machine learning, and data science. Generative AI has been making waves, and companies are eager to incorporate AI into all facets of their business processes. With the goal to lead in Business AI, SAP offers an extensive range of solutions from standalone AI services targeting a technical audience to embedded AI use cases in standard offerings, targeting users in the business departments.
While it makes perfect sense to provide different solutions, depending on the user group, business purpose or use case, this variety of options can be really confusing to someone who is new to the topic. But why do we offer so many AI options within the SAP portfolio? The answer is simple, sometimes it takes more than one superhero to save the world. In this blog post we want to share our personal experience and insights. We hope this analogy of AI infusing superhero powers to your business will help you better understand the breadth and depth of the SAP portfolio.
We all know from Sci-Fi and action movies, there are many superheroes with different skills and competencies living in different universes. Sometimes they face challenges on their own, or together as a team, depending on the task and goal. Let’s start with our embedded superheroes in the cloud.
In the SAP world, one of our embedded superheroes, who makes our daily life easier, is called Joule. Joule is a generative AI assistant and revolutionizes how customers can interact with their SAP business systems. For us there is no doubt, Joule’s impact will be huge. Not only for the business users but also for developers and how they can create clean-core compliant SAP extensions with i.e. SAP Build Code. Of course, there is not only a generative AI superhero in our embedded universe, but also classical AI use cases directly embedded into the business process. For example, there are predefined predictive use cases, like the Cash App, which can be embedded into the process in SAP S/4HANA Cloud. If the use case fits the customer’s needs, they train the predefined models with their own data and are ready to use the advanced technology in their application. So, customers can train the embedded superhero with specific skills required in their situation. An extensive summary of all the available embedded use cases for the different industries is available under the following link.
To make these AI use cases possible we need a strong (technical) foundation for our superheroes. This is where they gain all their superhero powers. Often, customers want to embed their own AI use cases into their business processes. Hence, SAP offers the AI Foundation powered by the Business Technology Platform. The AI Foundation consists i.e. of prepared AI Business services, HANA Machine Learning & Vector Database, SAC Smart Predict, AI Core & Launchpad and a capability of AI Core called the Generative AI Hub.
Let’s start with our SAP AI Business Services consisting of the Document Information Extraction service, Data Attribute Recommendation service and Personalized Recommendation service. For example, through the Document Information Extraction service customers can extract entities from unstructured data and integrate them into the process. This service has become very flexible by adding generative AI capabilities. I encourage you to give this a try in the BTP Trial edition following these tutorials. The Data Attribute Recommendation service can be a life saver if there are missing values in the dataset. Nothing can be more frustrating! Customers can train this service on their own data following this developer tutorial.
What is a superhero without a mission? Business context is key to bring an AI model into production. Even the most popular data science process (CRISP-DM) starts with the business understanding. In HANA Cloud and SAP Datasphere, customers can keep the business context of their data. Even better through HANA Machine Learning SAP brought the algorithms to the data. Classification, regression, time series analysis, clustering, and many more statistical methods are available directly where the SAP data resides. If you are curious, give this hands-on tutorial a try. Also, the business user can use a variety of these algorithms through Smart Predict in SAP Analytics Cloud. SAP Analytics Cloud is our self-service tool for business intelligence, predictive analytics, and planning. Further, the planer can even combine the predictive and planning capabilities. In addition, Just Ask enables the user to query their data with natural language powered by AI. Stay tuned because there is more to come, link.
AI Core and AI Launchpad enable the developer and ML engineers to deploy their own models inside of the Business Technology Platform and integrate them with other SAP solutions. If you want to create your first AI project using AI Core, have a look at the following developer tutorial. Sometimes our local superheroes need help from other universes, therefore you can access different Large Language Models (LLMs) through the Generative AI Hub. Since the general knowledge of these models is limited to a certain point in time, these superhero’s might not understand the language of your business universe. To help here, SAP has announced a Vector Engine as part of HANA Cloud. With the Vector Engine, customers can embed company-specific information into the knowledge of LLMs and can use the power of generative AI in the specific corporate context. If you want to dig deeper, have a look at the following blog post.
In addition, there are many more partnerships which complement the SAP portfolio. Other vendors offer very powerful data science solutions, and the integration can support data scientists to bring more AI models into production. For example, have a look at the following reference architecture.
Do you face any challenges that require an AI superhero to solve? Let us know!
We introduced our superheroes to you and how different user groups are enabled to combine AI with their business. This is our vision of saving the business world.
I want to thank Holger Seubert and Prof. Dr. Sarah Detzler for their support in writing this blog post!
Best wishes
Yannick
