Tech Events in Heidelberg

When a language model is wrong, it rarely sounds like it. Bridging the gap between statistical rigour and the fluid, high-dimensional outputs of modern AI is one of the central challenges for anyone deploying these systems where errors have real consequences.
We are excited to welcome Lynton Ardizzone, PhD in machine learning from Heidelberg University and independent ML/AI consultant, to our joint heidelberg.ai / NCT Data Science Seminar on 24th June at 5 PM.
In this in-person event, he will present a tour of uncertainty quantification for large language models, examining why classical UQ assumptions break down for these systems and what state-of-the-art approaches are doing about it.
We look forward to your participation in this seminar, where you will come away with a clearer sense of how far current methods get us, and what it would actually mean for a language model to know what it doesn't know.

For more info go to:
https://heidelberg.ai/2026/06/24/ardizzone.html

After the event, we will also upload a video recording to our YouTube channel:
https://www.youtube.com/channel/UCfHWBneOsb7SfOxJepnMQKA

Agentic Afterhours Volume 1

We are excited to invite you to our first volume of the Agentic Afterhours, powered by paretos.

Agentic Afterhours is a hands-on evening meetup for builders, founders, and practitioners exploring agentic AI — the systems and autonomous agents reshaping how software gets built and work gets done. As the first agentic AI meetup series in the Rhein-Neckar region, hosting it in Heidelberg puts this fast-moving field on the local map and gives the area's growing tech and research community a dedicated place to connect, share, and shape what comes next.

Our two speakers of the evening are Leon Ruddat, Founder of Strictly Boring, and Jan Hauer, Senior Director at exxeta.

Join us for an inspiring evening full of expert talks, networking, and great conversations!

Sign up now and secure your spot!

For updates please check below!

Our Agenda

18:00 Doors open
18:30 Welcome from the hosts

18:45 Leon Ruddat (Founder, Strictly Boring, former Snocks) - How does AI make money at SNOCKS?

19:15 Break: Networking with pizza and beverages

20:00 Jan Hauer (Senior Director | Member TI Board | Division Lead at exxeta) - You can’t modernize what nobody understands:
How AI agents helped make 1.6 million lines of legacy code manageable

20:30 Lightning Talks
20:45 Networking with snacks and beverages
21:30 End

We are looking forward to meeting you!

---

Join remotely/online:
Please check again later

---

Our speakers & topics:

Leon Ruddat (Founder at StrictlyBoring):

How AI Makes Money at Snocks

Leon knows both sides: For three years, as Head of AI at SNOCKS, he made sure that AI didn't stay a buzzword on a slide but actually made money — today, with his agency Strictly Boring, he builds exactly these kinds of systems for others. In this talk, he takes attendees behind the scenes: which AI solutions the team at SNOCKS really built — from automated newsletter translation via Klaviyo, to Facebook ads automation, all the way to automated Amazon advertising — what they cost, what they delivered, and where they also fell flat on their faces.

Jan Hauer (Senior Director | Member TI Board | Division Lead at exxeta)

You can't modernize what no one understands:
How AI agents helped make 1.6 million lines of legacy code manageable
Many companies face the same problem: the core system has been running essentially reliably for decades, but no one understands it anymore. Together with Real Garant, we are modernizing a central insurance system with 1.6 million lines of RPG code on an IBM AS/400 platform — with no documentation, no tests, and using a technology that is barely mastered by anyone today.
Instead of treating the migration as a translation problem, we deploy specialized AI agents to reconstruct business logic, dependencies, and system knowledge, and to make them usable for developers and then for the business departments as well.
In the talk, we show how a hybrid team of 25 developers and AI agents carried out a migration to Java and into a cloud-native target architecture on AWS/OpenShift in less than 18 months.
The key insight: the greatest leverage of AI in legacy projects lies not in generating code, but in making systems understandable again — and therefore modernizable.