Agenda

Since August 2023, Dyn has been the new home for handball, basketball, volleyball, table tennis and field hockey. In the presentation, Johannes and Felix give an insight into how Dyn built a BI architecture for a streaming service with more than 2,000 live games per season within a few months.

Automation platforms are becoming more metadata-driven, utilizing metadata to orchestrate and automate data management processes. Automation tools can intelligently automate tasks such as data integration, transformation, and governance by leveraging metadata to comprehend the structure, relationships, and lineage of data assets. VaultSpeed is a pioneer in this evolution. We provide the only automation tool that blends business understanding with metadata from all sources to suggest a comprehensive company-wide data model and automates the transformation and integration of data to populate that model. The underlying magic lies in our built-in integration templates, which generate ETL and DDL instructions for various sources and technologies, encompassing 2220 unique combinations. VaultSpeed also champions the ability to allow data engineers to code custom business logic automation, such as KPIs, and apply that across all data presentations in the information layer. We're thrilled to offer you a sneak peek of our latest advancements in Automated Data Transformation. Our team is currently focused on creating and testing features that will make it easier for users to perform customized data transformations, both upstream and downstream. We're diligently working on an intuitive, graphical drag-and-drop interface that will enhance accessibility to data automation for every data team. This will enable them to deliver consistency throughout the organization in alignment with their business objectives and data governance policies.

The leaders in tomorrow's economies will be data + ai companies. Today, the Data Lakehouse is an established best practice as it brought the performance and reliability of Data Warehouses to the elasticity and cost effectiveness of Data Lakes. But what about AI? Tomorrow, the Data Intelligence paradigm embeds AI in every layer and all capabilities of the modern data platform to transform businesses around the world. During this breakout session we'll take a look behind the scenes and what to expect in the future.

AI or ML processes often require data to be pulled from the data store, processed, and then the results/predictions stored back in the data store. This introduces a certain level of complexity and should be avoided. We show you how an AI/ML model can be executed directly at the location of your data without the need to move it beforehand - an advantage for keeping complexity and costs low while maintaining high quality.

We explore an innovative approach to mascot creation that combines AI agents and fine-tuned foundation models to uphold brand standards. While traditional foundation models offer a solid starting point, they often lack the specificity required for brand-centric content. By integrating AI agents alongside Stable Diffusion technology and a user-friendly interface (UI), we provide insight into a solution that enables creative employees to generate our beloved mascot, the POINTEE, tailored to PAYBACK's unique identity. Join us as we explore how this synergy between AI agents and foundation models can revamp content generation, but also where it reaches its practical boundaries.

Bitte bringen Sie Ihren eigenen Laptop für die praktische Übung mit. Keine AWS-Vorkenntnisse erforderlich, Grundkenntnisse in Programmiersprachen sind hilfreich. In diesem Workshop werden die Teilnehmer:innen einen generativen KI-Prototypen auf Basis von Amazon Bedrock entwickeln. Durch eine Reihe von praktischen Übungen, die sowohl für Anfänger als auch Fortgeschrittene geeignet sind, werden die Teilnehmer:innen lernen, die Fähigkeiten von Large Lanugage Models (LLMs) einzusetzen und mittels Amazon Bedrock Geschäftsprozesse zu verbessern. Amazon Bedrock ist ein vollständig verwalteter Dienst, der eine Auswahl hochleistungsfähiger Foundational Models (FMs) von führenden KI-Unternehmen wie AI21 Labs, Anthropic, Cohere, Meta, Mistral AI, Stability AI und Amazon über eine API-Schnittstelle bereitstellt um generative KI-Anwendungen sicher, gemäß aktueller Datenschutzverordnungen und auf Basis von ‘Responsible AI’ zu entwickeln und betreiben.

AI is already reshaping every industry, by augmenting our capabilities, transforming work, and shaping a new world of possibilities. In this session we will talk about how Microsoft's Cloud offering can help you power your AI transformation. Right from Copilots that help you become more productive to advanced LLMs and SLMs, we will look at how you can lead the Era of AI. We will also talk about building modern Data Platforms that act as the foundation for your AI initiatives. With Microsoft Fabric for Analytics & Microsoft Purview for Governance & Security, we got the entire Data Estate covered.

MediaMarktSaturn needs to enable data analytics at scale to meet the increasing market requirements. The talk illustrates the journey from the centralized analytics of the BICC, via a complete decentralized Data Mesh approach, towards a hybrid setup in the new Analytical Platform. We look at pros & cons of each concept based on first-hand experience. From that, we conclude how to achieve scalable analytics, while also aiming for company-wide adoption. Regarding the latter, organizational structures, analytical maturity, and legal requirements can be more challenging than the technology itself.

Learn how WhereScape improves Microsoft Fabric. Learn about streamlined migration, low-code integration, and improved metadata management that increases data management efficiency. Experience the synergy between WhereScape and Microsoft Fabric for advanced data technology.

The sports retail industry increasingly relies on leveraging data as part of day-to-day operations. With their newly modernized emnosSPORTS platform, emnos has witnessed true transformation based on real-time collaboration between sports retailers and sports brands. To achieve this, they require intuitive, flexible data analytics and visualization for multiple user types and organizations - a familiar challenge for emnos given their 15 years experience in the sports retail and food industries. Yellowfin and emnos have partnered to accelerate time to value for emnos’ clients, the sports retailers as well as the sports brands. Yellowfin provides the embedded analytics platform which enables emnos to deliver sophisticated sales, customer & stock KPIs, and forecasts to the sports industry - ultimately maximizing sales, reducing planning inefficiencies, and ensuring the right product at the right time in the right place to win with the consumers.

Truly complete data vault automation means that data vault expertise is no longer required. With this vision in mind, Reinhard and Marco developed Yotilla. Yotilla's innovative approach fully automates Data Vault data modeling and enables the definition of business logic without knowledge of the Data Vault data model. In the Breaktalk, we will present the way Yotilla works and show how Yotilla automates all changes to the data warehouse so that you can keep pace with changing or new business requirements at the same time. Using a customer example, we will show that Yotilla both significantly reduces the costs of creating and developing the data warehouse and generates benefits much faster.

Make Data and AI a team sport! Bring AI to all your data (structured and unstructured) with a simple approach. Understand how to activate GenAI powered data insights. Learn what features on Google’s Data Cloud help you to set up a mature Data and AI ecosystem.

Organisations try to federate the access to meaningful data and thereby generate value in a more efficient way. Usually technical roles are involved in the process of transforming raw data into consumable data products. By providing alternative entry points for less technical people to participate in the transformation process and starting to self-serve their needs for meaningful data products, quicker time-to-value can be achieved.

In the realm of AI Assurance and ensuring trust through agile data management, the emergence of RAG (Retrieval augmented generation) represents a pivotal advancement. RAG's integration of retrieval and generation techniques fosters richer and more contextually relevant responses, enhancing transparency and accountability in AI systems. By leveraging pre-existing knowledge alongside generative capabilities, RAG promotes robust data governance and bias mitigation, crucial components in fostering trustworthiness. In this session you will learn about the crucial role that effective Data Management plays in conjunction with Gen AI and how the Denodo Platform can enhance Gen AI by providing greater context.

The users of a modern data & AI platform can be very different with their individual needs in terms of development productivity, performance, user-friendliness and, of course, the use of the latest ML & AI technologies. In this talk, we will take a closer look at three user types: DATA PRODUCT BUILDER - as a technologist, is interested in quickly converting technical requirements into intuitively usable data products. Access to all relevant data, high-performance pipelines, data quality and professional presentation/management of data products are important success factors here. CITIZEN DATA SCIENTIST - as a business user, prefers to analyze specific questions "at the speed of thought" and appreciates the ease of use of data products and ML/AI functions. ML AND APPLICATION ENGINEER - helps to put an AI experiment/MVP into production, e.g. in the form of an intelligent, scalable app. The focus here is on the ability to use suitable software/frameworks (on a container basis), apply MLOps practices and operate the entire solution securely. How well does the Snowflake perspective fit your company's data strategy? The Snowflake team on site is looking forward to numerous and lively discussions after this talk!

OTTO is transforming itself into a platform and is no longer just a retailer, but also a marketplace provider. The number of marketplace partners is growing continuously. This offers many opportunities - but also some challenges. In the presentation, you will learn how we want to make the support of our partners more efficient in the future with the help of a GenAI-based chatbot and why this has given us a whole zoo of colourful little animals for AI optimization.

At LAYA, we have opted for an unconventional approach: push instead of pull! In this talk, we'll tell you why and how we did it. With our Azure Data Factory, Snowflake and dbt-based architecture, we'll show you how we overcame the challenges and what we learned from them.

Join us to explore the World of Data Vault Modeling and discover the keys to successful agile EDW deployments. We will answer the questions: What is Data Vault Modeling? How is it different from other techniques? Why do we use it? And, how can we apply DV successfully? This is a dynamic and interactive session where we will pursue ideas, discuss patterns and take the first steps into modeling a Data Vault. And as a bonus you will meet other data modelers and learn something new.

Artificial Intelligence (AI) is currently a central topic and is revolutionizing industries with a dynamic reminiscent of the rapid rise of the internet. Algorithms such as regressions, collaborative filtering, or factorization have been used for years to generate recommendations. But what possibilities do the latest generative technologies open up, and how can we further harness their potentials? In this presentation, you will be introduced to effective use cases showcasing the innovative capabilities of Bloomreach AI "Loomi" and gain insight into how these new technologies are transforming into inspiring opportunities.

In the Provocative Debate Forum, we invite our audience to put forward bold and possibly provocative theses for discussion in order to challenge our speakers on stage. The speakers are given 100 seconds to respond to each thesis. This promises to be an exciting and intensive discussion of important topics and will certainly lead to interesting insights. Participants: Andreas Wiener, Managing Director of BI or DIE, Pia Wolf, Founder & Managing Director of agrantis, Simon Faatz-Riccò, Team Lead Data Engineering at RND, Friedrich Staufenbiel, Managing Director & Head of Tech at Berlin Bytes, and Sebastian Amtage, Founder & Managing Director of b.telligent.

AI in the tax department is here to stay. I use three specific examples to show the added value that AI can already deliver in tax departments today.

Our unit has the opportunity to move to the AWS Cloud with its platforms and data. This is a positive perspective and a big challenge at the same time. Alongside the technical transformation, we are trying to re-skill the entire workforce and carry out a cultural transformation. The main challenges we are facing are: • Dealing with uncertainty • Getting buy-in from employees for the move to the cloud • Implementing the re-skilling • Giving people a sense of belonging and psychological security in a complex environment Learn more at our talk on how we dealt with these challenges and which measures worked best! We look forward to seeing you there!

In today's digital landscape, organizations are increasingly recognizing the benefits of migrating their data warehousing infrastructure to the cloud. This transition offers unparalleled scalability, flexibility, and enables businesses to unlock the true potential of their data assets. However, successful cloud migration requires careful planning, an agile mind set, and a brave mentality for the challenges and opportunities that lie ahead. This conference talk aims to provide a comprehensive overview of the P7S1 cloud migration in the context of data warehousing. We will delve into the key considerations, good practices, and pitfalls that organizations must navigate to ensure a seamless and successful migration.

Hear the compelling story of our attempt to move a 15-20 year old legacy data warehouse from on-premise to the cloud and why it didn't work out. Despite years of investment and effort, the project failed. We’ll break down the reasons, from technical issues to strategic missteps, and share the new approach we’re using to make sure we succeed this time around. Learn how to avoid common pitfalls and explore better alternatives to traditional cloud migration.

The approach of a central Data Warehouse for everything can be seen as a failure – too expensive, too slow, too cumbersome. Nonetheless, how can a holistic data-based view of the business be achieved? How to deal with data silos und various business processes after acquisitions? How to balance data platform building versus providing specific answers to business questions? Dr. Clemens Schäffner will discuss these questions and show how Gerresheimer developed a suitable data strategy, what specific implementation steps were taken and how the long-term vision is approached step by step.

Complete transparency across all business processes and their data representation has long been a dream of many. However, in the manufacturing industry in particular, this involves a high level of investment in order to capture the relevant data within production. This is followed by centralization, standardization and integration with other “classic” data sources. Fränkische IP has risen to this challenge and now operates a cloud platform that makes it possible to analyze data from incoming goods through production to delivery in an integrated manner. A traceability portal provides Fränkische IP's customers with extensive analysis options for their QA processes and further planning of production capacities.

Nowadays, “AI” is a buzzword for everyone (including Senior Management). Many Data & Analytics teams are exploring how to leverage the new potentials of Generative AI for their work processes. The speech explains the common dangers of Gen AI, the impact due to regulations frameworks and will concentrate on how DKV Mobility is creating their AI framework to prioritize on business value.

Football stands as a cherished global phenomenon, and the Bundesliga is dedicated to continually uplifting fan engagement to unprecedented levels. Our ambition is to “BECOME THE MOST FAN-CENTRIC LEAGUE IN THE WORLD”. Through harnessing cutting-edge technologies and pioneering strategies, we are empowered to deliver personalized content and immersive experiences for all.

Carsten Schweiger, DWH Engineer and Agile Coach, shows how lengthy and inflated DWH projects are becoming a relic of the past. A model-driven data vault approach and short feedback cycles bring business and IT together. Automation and standardization in the development process drastically reduce costs and effort. Welcome to a world in which time-to-market is no longer an obstacle, but a competitive advantage.

Based on her extensive experience, Annabell Whitney provides exclusive insights into the rock star life and the secrets of their success. She draws surprising parallels between the two worlds and inspires participants to ignite their own rock star DNA in a captivating and musical way. This keynote releases unimagined energy to spark motivation and enthusiasm - for personal success.

The constantly growing number of multichannel marketing campaigns increases the complexity of cross-departmental collaboration for planning and implementing measures. In the past, Excel-based, decentralized planning led to a great deal of manual effort and inconsistencies. The introduction of a cloud-based marketing resource management (MRM) solution creates a single point of truth. Thanks to the integrated workflows, the parties involved can work together much more efficiently.

Introducing agile working methods in a single team is relatively simple. Implementing it for several teams in a department such as DHW, Reporting or BI is much more complex. At S-Kreditpartner (SKP), we have adapted scaling Scrum methods for ourselves and introduced an agile way of working based on Scrum and Kanban in the Data Management department. With success! This experience report describes our journey from the start with several, independently operating development teams to a scaling organizational unit with integrated teams. We will use examples to describe the challenges and experiences of this journey and the lessons we have learned along the way.

In thermal waste treatment, extensive operational data is generated. The challenge lies in the complex capture and analysis of this data while maintaining data integrity and cybersecurity to make it usable for quick and informed decisions. Our goal is to increase efficiency and reliability by combining data processing and OEM process knowledge in combustion and flue gas cleaning. Our solution leverages edge-to-cloud data transmission and analysis on the Stackable platform in the IONOS Cloud.

Metadata as "data about data" is intended to help find, understand, and trust data. However, the topic is often approached purely technically, neglecting the substantive significance of the data. We discuss what is needed for a human user to gain insight and understanding for working with data from metadata, and the implications this has at Commerzbank for the establishment of a Data Governance Program and the use of tools.