Discover how SAP enables enterprises to benefit from trustworthy business AI by automating processes, improving decision-making, and enhancing customer experiences. You will learn how recent advances of foundation models – accelerated by the big-ChatGPT-bang – are transforming business applications and the way users intact with machines. Johannes will showcase common AI scenarios illustrating the transformative power of these models and give an outlook on what research topics will be key for the intelligence layer of modern enterprises.

Presenter:

Johannes Hoffart is heading the CTO AI office at SAP, a group of technology experts and scientists driving SAP’s AI technology strategy and developing core AI technology assets. As CTO of the AI unit, Johannes is ensuring SAP products incorporate AI innovations at the right time and pace. Before joining SAP in 2021, Johannes has led an AI research group on NLP and Knowledge Graphs at Goldman Sachs and co-founded a spin-off from the Max Planck Institute for Informatics with the goal of enabling businesses to tap into their knowledge hidden in text.

Bckground Info:

The internet is multilingual, and thus inherently are LLMs. This talk will try to give some food for thought on multilinguality and why evaluating multiple languages in their respective semantic contexts matters.

Presenter:

Lauritz Brandt is a Data Scientist working with SAP on Machine Translation. His current focus is bridging the gap between new research and business requirements by scaling Machine Learning advancements to enterprise level. Lauritz started with SAP directly after graduating from Heidelberg University with a Master's Degree in Computational Linguistics in 2018, and has worked on NLP-related topics since.

LLMs (like ChatGpt or GPT4) have been heavily in the focus in recent months. The talk will give a general overview of how we arrived where we are and will attempt to provide a stimulus of discussion on applications and issues.

Presenter:

Georg Groh’s main fields of research are social computing and applied natural language processing. He is currently working as an adjunct professor at TUM School of CIT and head of the Social Computing research group. Current research activities include (among others) explainable and ethical AI with a focus on NLP, NLP techniques for simplified language applications, reasoning with LLMs, NLP techniques for topic landscapes, and general machine learning based NLP. Teaching includes several lectures and lab courses on NLP and social computing. 

Business process compliance means to assess whether or not business processes comply to relevant compliance requirements such as laws, regulations, or guidelines. The input for this task is manifold, ranging from business process models, textual process descriptions, and event logs at the process side to mostly textual input at the compliance requirement side. The latter can be quite complex, e.g., the EU adopts more than thousand legal acts per year, and can change quite frequently, for example, bank regulations change every 12 minutes. This talk highlights how AI-based methods can support the task of business process compliance ranging from the extraction of compliance-relevant information from text, over compliance verification of process event logs over textual compliance requirements, to conversational process modeling using chatbots.

Presenter:

Stefanie Rinderle-Ma is a full professor at the Technical University of Munich, Germany, and holds the Chair of Information Systems and Business Process Management. Her research interests focus on process-oriented information systems, flexible and distributed process technologies, compliance management, as well as production and process intelligence. The overarching goal of her research is to enable and accelerate digitalization and automation through processes and at the same time keep the human in the loop. Application areas comprise manufacturing, transportation and logistics, as well as medicine.

With the increased complexity of supply chains and the continued digitization of business processes, the amount of data that is available to help steer them is growing immensely. With the use of AI techniques, this data can be used to take more informed decisions during the supply chain planning and execution. In the presentation we will give examples of how SAP offers such enhanced AI-based decision-making in its different software products for supply chain.

Presenter:

Dr. Uta Lösch is heading one of the Data Science teams in CloudERP AI Incubation. Her team supports the different product units in the CloudeERP solution area with their Machine Learning use cases from ideation to productization. The focus is on topics in Digital Supply Chain, Asset Management, and Sustainability. She obtained a Ph.D. in Machine Learning and Semantic Web from Karlsruhe Institute of Technology prior to joining SAP as a data scientist in the development organization for Predictive Maintenance in 2012.

We as IT department within SAP are responsible for operating the entire IT systems which are used to manage our processes and economic transactions at SAP. We continuously enable SAP to be an intelligent, sustainable and data-driven enterprise. Among other things, we have been delivering hundreds of AI solutions over the years and have recently started the initiative to build a digital AI Factory to scale up our delivery of AI use-cases, streamline our processes and increase our overall efficiency. In this session, you will learn about the purpose, key motivations and components of the AI factory that we envision.

Presenter:

Darwin is heading the AI Center-of-Excellence team within SAP IT, based out of Walldorf/Germany. In this role, he is responsible for the delivery of AI platform and solutions across various LoBs. He has 15+ years of experience in data and analytics area and assumed various roles from developer, (solution / enterprise) architect, (project / program) manager and market unit lead across his career. He holds a master’s degree in Computer Science at RWTH Aachen University and is currently enrolled in Executive MBA program at IESE Business School in Munich.

This project undertakes an exploration into the utilization of reinforcement learning techniques as part of the decision-making process for extubating patients on mechanical ventilation in Intensive Care Units (ICUs). The primary objective is to implement innovative reinforcement learning strategies to aid clinicians in making individualized and secure extubation decisions. This involves addressing several specific challenges such as managing data imbalances and adhering to clinical safety protocols in practice. The project will leverage concepts from constrained reinforcement learning and offline reinforcement learning to navigate these challenges.

Presenter:

Prof. Dr. Jingui Xie is Professor of Business Analytics in the School of Management, Technical University of Munich. His research interests include business analytics, reinforcement learning, optimization with prediction, queueing theory, and healthcare management.