AI Meetup @SRH Gründerinstitut

Info: This event will be held in English.

The AI Meetup is a regular event that aims to promote dialogue and collaboration within the AI community in Heidelberg and beyond. The mission of this event series is to connect scientists, researchers, industry experts, students and alumni to expand knowledge and intensify dialogue. Whether beginner, advanced or expert - everyone is welcome here.

Focus in this session:
This series of events is organised in cooperation with HMS Analytical Software GmbH and takes place in the innovative premises of the SRH Gründerinstitut. Come along if you would like to discuss AI engineering and current research on universal AI and expand your knowledge.

Places are limited to 50 people - first come, first serve.

Details on the speakers and their topics:

Christoph Bergen

From data scientist to AI engineer - What has changed? What does it take? A consulting implementation perspective The data scientist was considered the most attractive job of the 21st century. With the advent of generative AI, a new role has emerged: the AI engineer. In many organisations, data scientists have taken on the tasks associated with this new position. However, the skills required for both roles are not completely interchangeable. While there is some overlap, the AI Engineer role requires additional, distinct competences. In this presentation, we will share our observations from the perspective of an AI implementation consultant. We will walk through our own path of transition and highlight the key differences between the roles of Data Scientist and AI Engineer. In addition, we will discuss the concept of Atomic Delivery Teams and how they can be structured effectively in AI projects. By examining different project categories, we will highlight further specialisation opportunities for AI engineers. The aim of this immersion is to create a better understanding of how the evolving AI landscape impacts roles, required skills and team structures in a practical consulting context.

Johannes Haux

Is the AI race over? Our bet on Europe’s return to the top at kausable

The global AI race has been dominated by massive models, vast datasets, and exorbitant training costs. But what if there’s a better way?

In this talk, we’ll take a look at how we can rethink the fundamentals of artificial intelligence, not as a matter of scaling bigger models but of building systems that truly understand, adapt, and reason. We'll explore emerging ideas like inference-time learning, agentic behavior, and synthetic priors, and how they challenge the assumptions behind today’s dominant approaches.The goal isn’t just to build more powerful models. It’s to make AI more human-aligned, capable of learning from experience, making decisions under uncertainty, and improving continuously in dynamic environments.

This is about moving beyond prediction toward understanding and imagining what AI could be if we started from first principles.