Day 1 Online Workshops

“It depends.”

It’s the classic software architect’s answer… but what does it actually depend on? Patterns like Clean and Hexagonal Architecture are great tools, but they are often applied dogmatically. When we automatically wrap simple features in endless layers of abstraction and domain interfaces, we introduce unnecessary distance between related concepts. Those extra layers are intended to decouple, but they can sometimes just make our code harder to navigate, test, and maintain.

In this talk, we’ll take a step back from “best practices” and look at the trade-offs behind architectural choices. Using clear Java code samples, we will explore the metrics that help guide these decisions: coupling, cohesion, and distance. Instead of looking for universally “clean” solutions, we’ll discuss how to evaluate what makes sense for your specific context.

Implementing streams? Adding WebSockets to communicate with your agents? That’s one of the major challenges in connecting your UI to your created agents. But what if there were a universal adapter so you didn’t have to build those connections yourself?

In this session, we will learn how to implement the AG‑UI protocol to easily connect UI apps with pre‑built agents. No matter what framework you used to build your agents, AG‑UI will handle the work for you!

AI agents are reshaping the future of work, and harnesses will be central to making that future effective. While vibe coding has brought AI‑assisted creation into the mainstream, professional developers are already looking beyond it toward more structured, reliable workflows.

The next step is to make AI workflows more deterministic, controllable, and effective. Harness architecture provides that layer of control—enabling teams to guide agents more precisely and consistently deliver higher‑quality outcomes.

In this talk, you’ll learn how to design and build your own harness using the new Zaumzeug (ZaZu) Harness SDK, and see how these principles can be put into practice

AI is already doing most of the work. Detecting flaky tests. Triaging bugs. Generating test scenarios. Writing incident summaries. So here’s the uncomfortable question: what exactly are you doing?

This session isn’t about whether AI will replace testers. That debate is boring. This is about what happens after AI takes over the routine work and whether you’re building the skills that actually matter now.

Through real implementations (and a few experiments that failed badly), we’ll look at where AI genuinely helps in quality engineering, where it quietly goes wrong, and why the humans who understand both will be the ones running the show.

Because AI doesn’t think. It predicts. And there’s a massive difference, especially when your production environment is on fire.

Choosing an LLM is easy when you’re comparing benchmark leaderboards. It’s much harder when you’re deciding which model to trust with real customer conversations, at scale, under cost constraints.

This talk walks through ChatGenie’s evaluation of five leading models — GPT-5.2, Claude Sonnet 4.5, Claude Haiku 4.5, Qwen 3 Next 80b, and Amazon Nova Pro v1 — across 102 real production scenarios pulled from live enterprise deployments. We’ll cover why the model that won on composite score wasn’t the one that got promoted to production Orchestrator, what “accuracy” actually means when a support agent has to handle ambiguity and multilingual queries, and a practical framework attendees can use to evaluate LLMs against their own production needs rather than someone else’s benchmark.

Ideal for engineers, AI leads, and product owners making build-vs-buy or model-selection decisions right now.

As software teams cross into the next horizon of AI-assisted development, cloud-native systems, automated pipelines, and faster release cycles, security must evolve beyond individual code fixes.

This session explores how teams can move from secure coding to secure software delivery by embedding security into design, development, testing, deployment, and continuous improvement. Participants will learn practical approaches to secure-by-design thinking, automated security checks, dependency management, secrets protection, AI-generated code risks, and building a culture where security becomes part of everyday engineering decisions.