LE Software Craft Community

We're back with another interesting topic, this time in english!

Abstract
Everyone's talking about AI as a tool for writing code faster. But that's just one way AI has entered your life as a developer — and arguably the least consequential for what you actually ship.

This talk is about how AI changed what we can and should build.

AI agents are rapidly becoming primary consumers of software. As humans delegate more tasks to autonomous agents, we need to think beyond UX to AX — the agent's experience. We'll explore what it takes to build software that agents can and will actually use, so your products stay relevant as user behavior shifts.

At the same time, publicly available models and LLM APIs have made it easier than ever to integrate AI-based capabilities into your own software. But these probabilistic components behave fundamentally differently from everything else in your system and come with pitfalls most developers only discover the hard way. We'll address how to avoid them and capitalize on the new opportunities AI unlocks.

No background in machine learning required. But you'll leave with a clearer picture of the decisions that will determine whether what you build survives — and thrives — in the age of AI.

Our speaker Dr. Franziska Horn:
Dr. Franziska Horn is a freelance AI & Software Design Consultant. She holds a PhD in Machine Learning from TU Berlin and combines deep academic expertise with 12+ years of hands-on experience as a data scientist and tech lead across startups and large enterprises.

Her focus is on AI that delivers real business value, not marketing fluff. She approaches ML projects with production in mind from day one, building solutions that go beyond prototypes and create lasting impact.

Pizza and Drinks are on codecentric, thanks a lot!

Agenda:
Presentation: 18.00
Pizza: 19.15

We're back with another interesting topic, this time in english!

Abstract
Event-sourcing allows the business to ask questions about your application's data that weren't thought of when the system was created, such as "how often are customers changing their shipping address?".

However, it requires a mind-shift from the well-known DDD Tactical Patterns (Aggregates and Repositories), but the benefits are more flexible systems, with a state-change trail that AI can leverage.

In this session, you'll learn how event-sourcing works at the code level without event-sourcing libraries getting in the way of deeper understanding. We'll walk through a codebase for a Concert Ticketing system and see how straightforward the implementation can be.

We start with "Event Modeling" the system, defining the commands (user actions) that generate them, the aggregates that make business decisions (and how to move to "deciders"), and the projections used to generate the user interface. We'll see how event-sourcing makes testing easier to write and understand.

We'll end by touching on the challenges to using event-sourcing, such as performance and schema evolution.

Our speaker Ted M. Young:
Ted M. Young (he/him) is a Java trainer, coding coach, speaker, and author. He’s been in software development for over 30 years, teaching and coding in Java since the 1.0 beta, and doing eXtreme Programming since 2000.

Ted has worked for eBay, Google, Apple, and Guidewire Software. He is now an independent educator, helping those new to the industry and experienced folks increase their joy in coding by making code more testable.

Ted’s favorite techniques are test-driven development, refactoring, domain-driven design, and testable architectures, such as Hexagonal Architecture and Event-Sourcing.
Pizza and Drinks are on codecentric, thanks a lot!

Agenda:
Presentation: 18.00
Pizza: 19.30