en.Wedoany.com Reported - The data2day 2026 conference will take place on October 7–8, 2026 in Cologne. The recently announced program reveals that this year's event will focus on data products, agentic artificial intelligence (Agentic AI), and data governance. Organized by iX and dpunkt.verlag since 2014, the conference is aimed at data scientists, data engineers, data analysts, software developers, and team leads. The two-day program of talks covers both introductory and advanced content, with a full-day workshop on October 6 featuring a "Data Contracts in Practice" course led by Simon Harrer of Entropy Data.
Generative AI and agentic AI have emerged as the central themes of this year's conference. Matthias Niehoff from codecentric will explore agentic data engineering, analyzing how agents build and repair data pipelines. Max Vieweg and Denise Hartmann from inovex will demonstrate how to systematically track and evaluate agent loops. Language models are also mentioned multiple times: Björn Buchhold from CID compares Text2SQL approaches, covering frontier models, open weights, and fine-tuning; Kaan Duran from E.ON Digital Technology explains how to use LLMs to improve machine learning features; and Tobias Otte from viadee proposes connecting LLMs with data platforms via a semantic layer. Jan Trienes and Matthias Richter from inovex will explain systematic evaluation methods for generative AI. Sohan Maheshwar from AuthZed will discuss how to prevent AI agents from accessing unauthorized data.
Data architecture, data quality, and data products form the second major theme. Sönke Liebau from Stackable positions the data catalog as a control instance within the open data lakehouse, while his colleague Florian Müller introduces modern data formats from microservices to ML models. Franz Wöllert from Heidelberger Druckmaschinen shares experiences using technologies like Iceberg; Andreas Buckenhofer from Adam Riese demonstrates the fusion of classic star schemas with cloud lakehouses. Matthias Niehoff from codecentric and Nicolas Renkamp from Merck KGaA jointly explore the reality of unbundled data stacks. On data quality, Danilo Brajovic from Fraunhofer IPA discusses data quality methods and blind spots for ML, and Felix Theodor from Otto Group one.O describes the evolution from manual test fixtures to AI-generated test suites. Daria Haselhoff and Markus Nutz from METRO Markets present stable data models and product construction, while Jochen Christ from Entropy Data advocates for trust through data contracts. Ramona Casasola-Greiner and Matthias Böck from pub.tech call for a product mindset over pure use-case logic. Nikolai Hofschulz from AUDI emphasizes the role of semantics in data development, and Marc Schubert from adorsys proposes a federated semantic layer to link KPIs. Evgeniya Alekseeva from Daikin Europe analyzes success and failure factors in data governance initiatives.
Industrial applications and legal aspects round out the agenda. Christian Haack and Björn Wendland from ControlExpert introduce conformal prediction for information extraction and cost estimation; Nikita Golovko from Siemens describes a domain-driven path to industrial AI; Eva Feigl from Forterro Deutschland Abas discusses metrics, A/B testing, and the role of product success; Konstantin Hopf and Deniz Oruç Çelik from TU Chemnitz reveal the tension between data science work and management expectations; and Joerg Heidrich from Heise Medien explains key legal compliance points for developing AI within enterprises.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
