Program
Time
Schedule
07:30 - 08:30
Registration and Expo
07:30 - 08:30
Registration and Expo
08:30 - 08:50
08:30 - 08:50
- Main Stage
08:50 - 09:45
08:50 - 09:45
- Main Stage
Opening Keynote
Software verification in the age of Artificial Intelligence
The AI tsunami is transforming every aspect of software engineering. What does it hold in store for the world’s software and for the profession itself? Beyond the buzz, can “vibe coding” scale up to the production of the high-quality systems the world increasingly requires? How do these evolutions affect testing and, more generally, software verification?
The talk will address these questions and, more generally, analyze what software development will look like in the new, AI-rich world. It will particulary focus on the need for a modern form of software verification, supported by advanced tools and combining the best of dynamic techniques (tests) and static ones (proofs), all supported by AI agents. It will also draw the lessons for software developers and testers who want to navigate the new landscape successfully.
Programmer, entrepreneur, author, language designer, researcher, software legal expert
- Main Stage
09:45 - 10:00
Coffee Break
09:45 - 10:00
Coffee Break
10:00 - 10:30
10:00 - 10:30
Global Head of Solution Management & Global Head of Digital Design & Experience Testing
Customer Zero – Perfecting the Experience Before It’s Airborne
What do aviation and diagnostics have in common? There’s no such thing as a harmless mistake. A misplaced alert, a confusing step, a subtle ergonomic miss or even a small bug can have drastic real-world consequences.
Continuous testing keeps those moments from reaching the people who rely on us the most: Patients. As AI and immersive technologies reshape our systems, we’ve built the CX Lab, a space where digital design and physical validation come together from day one. Using immersive simulations and digital experience twins, we test the entire user journey before a single component is built.
This approach culminates in our “Customer Zero” Integration Lab, our ‘customer-before-the-customer’, for final, real-world validation. Join us to see how we define quality and pioneer change in a space where reliability is critical and customer Trust is everything.
- Main Stage
- Main Stage
- Main Stage
Test Automation at Scale with MCP and Playwright
The Model Context Protocol (MCP) has taken the world by storm as a standard for integrating AI systems with external tools and workflows. In this talk, we show how MCP can be combined with Playwright to enable intelligent, automated end-to-end testing. By connecting an MCP Playwright server to AI agents, we unlock the ability for LLMs to generate, execute, and adapt tests reducing manual effort and accelerating feedback loops. Attendees will learn how this approach bridges AI reasoning with browser automation and see real-world examples of AI-native testing pipelines.
- Room 3
Test Automation at Scale with MCP and Playwright
The Model Context Protocol (MCP) has taken the world by storm as a standard for integrating AI systems with external tools and workflows. In this talk, we show how MCP can be combined with Playwright to enable intelligent, automated end-to-end testing. By connecting an MCP Playwright server to AI agents, we unlock the ability for LLMs to generate, execute, and adapt tests reducing manual effort and accelerating feedback loops. Attendees will learn how this approach bridges AI reasoning with browser automation and see real-world examples of AI-native testing pipelines.
- Room 3
- Main Stage
Good AI Testing Strategy / Bad AI Testing Strategy. The difference and why it matters
What makes one AI testing strategy succeed where another fails? Is it the technology, the people, or the funding behind it or is it something deeper, something about how we think about strategy itself? In today’s fast-moving landscape of AI integration, the distinction between good and bad strategies is not just a matter of preference. It is a matter of survival.
But what truly defines a good strategy in AI testing? And how do we recognise a bad one before it’s too late? Drawing inspiration from Richard Rumelt’s ideas, we will begin by clarifying what the word “strategy” really means, beyond the buzzwords and lofty goals. Many organisations mistake ambition for direction and confuse lists of actions or aspirations with coherent strategic thinking. The result is often what Rumelt calls a “bad strategy”: one that hides behind slogans, ignores the real problem, and substitutes wishful thinking for diagnosis.
So why do even smart, capable teams fall into this trap? What happens when companies chase “AI transformation” without first understanding the systems, data, and operational realities they are dealing with? And what are the consequences of testing becoming a vibes confirmation rather than a means of discovery?
In contrast, what does a good AI testing strategy look like when applied to a complex system? How can it align with enterprise objectives, foster accountability and build operational resilience? Most importantly, how can it balance the need for speed with the discipline of evidence-based verification, especially in a world where AI’s behaviour is probabilistic and constantly evolving?
Throughout the session, we will explore these questions through the layers of verification that define large-scale AI implementations. I will share concrete examples from practice that illustrate both the pitfalls of flawed strategies and the power of sound, principle-driven ones.
By the end, the challenge will be clear: is your AI testing strategy built to impress, or is it built to endure?
(The topic can be a part of either the AI Governance & Trust or the Human-centered AI Testing streams)
- Room 4
Good AI Testing Strategy / Bad AI Testing Strategy. The difference and why it matters
What makes one AI testing strategy succeed where another fails? Is it the technology, the people, or the funding behind it or is it something deeper, something about how we think about strategy itself? In today’s fast-moving landscape of AI integration, the distinction between good and bad strategies is not just a matter of preference. It is a matter of survival.
But what truly defines a good strategy in AI testing? And how do we recognise a bad one before it’s too late? Drawing inspiration from Richard Rumelt’s ideas, we will begin by clarifying what the word “strategy” really means, beyond the buzzwords and lofty goals. Many organisations mistake ambition for direction and confuse lists of actions or aspirations with coherent strategic thinking. The result is often what Rumelt calls a “bad strategy”: one that hides behind slogans, ignores the real problem, and substitutes wishful thinking for diagnosis.
So why do even smart, capable teams fall into this trap? What happens when companies chase “AI transformation” without first understanding the systems, data, and operational realities they are dealing with? And what are the consequences of testing becoming a vibes confirmation rather than a means of discovery?
In contrast, what does a good AI testing strategy look like when applied to a complex system? How can it align with enterprise objectives, foster accountability and build operational resilience? Most importantly, how can it balance the need for speed with the discipline of evidence-based verification, especially in a world where AI’s behaviour is probabilistic and constantly evolving?
Throughout the session, we will explore these questions through the layers of verification that define large-scale AI implementations. I will share concrete examples from practice that illustrate both the pitfalls of flawed strategies and the power of sound, principle-driven ones.
By the end, the challenge will be clear: is your AI testing strategy built to impress, or is it built to endure?
(The topic can be a part of either the AI Governance & Trust or the Human-centered AI Testing streams)
- Room 4
- Main Stage
Beyond Execution: Human Oversight and Trust in AI-Enhanced Testing
AI is becoming an integral part of modern testing, accelerating test creation, failure analysis, and continuous validation. As automation becomes enhanced by AI, testers face new challenges. How can we trust AI-generated results, and what remains the tester`s responsibility when intelligent tools take on more of the work?
This talk explores what happens beyond execution, when testers move from running scripts to designing systems that remain explainable, traceable, and trustworthy. Based on both experience and current research trends in augmented testing, it discusses how human oversight and data quality form the foundation for successful AI-driven automation. Through examples and a brief illustration of how oversight can be embedded into automated workflows, the session shows how testers can evolve into the guardians of responsible automation. Participants will gain a clear understanding of how AI can augment, not replace, human testers and will learn practical strategies to build trust, transparency, and accountability into modern testing practices.
- Room 5
Beyond Execution: Human Oversight and Trust in AI-Enhanced Testing
AI is becoming an integral part of modern testing, accelerating test creation, failure analysis, and continuous validation. As automation becomes enhanced by AI, testers face new challenges. How can we trust AI-generated results, and what remains the tester`s responsibility when intelligent tools take on more of the work?
This talk explores what happens beyond execution, when testers move from running scripts to designing systems that remain explainable, traceable, and trustworthy. Based on both experience and current research trends in augmented testing, it discusses how human oversight and data quality form the foundation for successful AI-driven automation. Through examples and a brief illustration of how oversight can be embedded into automated workflows, the session shows how testers can evolve into the guardians of responsible automation. Participants will gain a clear understanding of how AI can augment, not replace, human testers and will learn practical strategies to build trust, transparency, and accountability into modern testing practices.
- Room 5
- Main Stage
Von Testdaten zu Vertrauen – Wie KI Governance im Software-Lifecycle Realität wird
KI ist längst Teil moderner Test- und Release-Prozesse – aber Vertrauen entsteht nicht durch Automatisierung allein. Unternehmen kämpfen mit Datenfragmentierung, intransparenten Entscheidungen und fehlender Nachvollziehbarkeit in ihren KI-gestützten Testsystemen. Genau hier zeigt sich die wahre Herausforderung: Wie stellen wir sicher, dass KI-gestützte Qualitätssicherung erklärbar, überprüfbar und auditierbar bleibt?
In diesem Vortrag geht es um die praktische Umsetzung von KI-Governance im Test- und Release-Kontext. Anhand realer Erfahrungen beleuchten wir, wie sich Feedback-Schleifen zwischen Entwicklung, Test und Betrieb automatisiert, aber kontrolliert schließen lassen – von der Testoptimierung über Release-Orchestrierung bis zur kontinuierlichen Risikobewertung. Ziel ist es, nicht mehr nur Software auszuliefern, sondern Vertrauen in die Entscheidungen der KI-Systeme aufzubauen.
Ein CTO-Blick auf das Spannungsfeld zwischen Geschwindigkeit, Compliance und erklärbarer Intelligenz.
- Room 6
Von Testdaten zu Vertrauen – Wie KI Governance im Software-Lifecycle Realität wird
KI ist längst Teil moderner Test- und Release-Prozesse – aber Vertrauen entsteht nicht durch Automatisierung allein. Unternehmen kämpfen mit Datenfragmentierung, intransparenten Entscheidungen und fehlender Nachvollziehbarkeit in ihren KI-gestützten Testsystemen. Genau hier zeigt sich die wahre Herausforderung: Wie stellen wir sicher, dass KI-gestützte Qualitätssicherung erklärbar, überprüfbar und auditierbar bleibt?
In diesem Vortrag geht es um die praktische Umsetzung von KI-Governance im Test- und Release-Kontext. Anhand realer Erfahrungen beleuchten wir, wie sich Feedback-Schleifen zwischen Entwicklung, Test und Betrieb automatisiert, aber kontrolliert schließen lassen – von der Testoptimierung über Release-Orchestrierung bis zur kontinuierlichen Risikobewertung. Ziel ist es, nicht mehr nur Software auszuliefern, sondern Vertrauen in die Entscheidungen der KI-Systeme aufzubauen.
Ein CTO-Blick auf das Spannungsfeld zwischen Geschwindigkeit, Compliance und erklärbarer Intelligenz.
- Room 6
10:00 - 16:00
10:00 - 16:00
Tool Playground
The Tool Playground is an interactive space where attendees can discover and experiment with a wide range of modern testing tools. Participants get the chance to try out tools in practical scenarios, deepen their skills, and gain first-hand experience in a collaborative, hands-on environment.
For sponsors, it offers a unique opportunity to present their solutions in action, allowing attendees to engage with and understand their capabilities without any sales pressure. This experience highlights the real-world applications of sponsor tools, increases visibility among a highly interested testing audience, and fosters meaningful interactions in an engaging and educational setting.
- Bar2020
10:30 - 10:45
Coffee Break
10:30 - 10:45
Coffee Break
10:45 - 11:15
10:45 - 11:15
Building an AI-supported Testing Ecosystem in Insurance
Discover how a Quality Engineering team in the insurance sector applies AI to set the grounds of an AI-supported Testing Ecosystem:
Optimizing the timelines on complex test data creation. Re-discover and challenge their E2E test coverage by combining data from test management system, test code base and incident management. Enhance pipeline stability by making AI analysis on failures. A candid story about the challenges, insights, and impact of delivering tangible results with the support of AI — without buzzwords.
- Main Stage
- Room 3
- Main Stage
Agentic testing in banking: From hype to governed practice
Agentic AI is moving from experimentation into production-grade testing, even in highly regulated banking environments. This session shows how agentic testing is already being applied in practice, building on LGT’s award-winning work with Infometis and the ongoing evolution of the Avalon Agent.
The talk demonstrates how agents use context grounding and Retrieval-Augmented Generation (RAG) over bank-internal test artefacts and systems, while Model Context Protocol (MCP) servers expose deterministic, auditable tools for agents to interact with enterprise systems in a controlled way. Rather than another AI vision talk, this session focuses on what already works, where the limits are, and how to operationalize agentic testing in banking.
- Room 3
Agentic testing in banking: From hype to governed practice
Agentic AI is moving from experimentation into production-grade testing, even in highly regulated banking environments. This session shows how agentic testing is already being applied in practice, building on LGT’s award-winning work with Infometis and the ongoing evolution of the Avalon Agent.
The talk demonstrates how agents use context grounding and Retrieval-Augmented Generation (RAG) over bank-internal test artefacts and systems, while Model Context Protocol (MCP) servers expose deterministic, auditable tools for agents to interact with enterprise systems in a controlled way. Rather than another AI vision talk, this session focuses on what already works, where the limits are, and how to operationalize agentic testing in banking.
- Room 3
Dude, where’s my Test Case? Wie LLMs die Testfall-Erstellung bei Migros Online revolutionieren
Was tun, wenn das Prozessdokument dicker ist als ein Roman, die Anforderungen massiv sind und die Zeit drängt? Man lässt die KI ran. Begleiten Sie uns auf die Reise, wie für das Warehouse Management System von Migros Online eine LLM-basierte UAT-Generierung implementiert wurde. Wir teilen unsere Learnings aus drei Entwicklungsphasen, diskutieren technische Stolpersteine und zeigen, wie aus einer komplexen Herausforderung ein Tool wurde, mit dem das Team seine Testfälle nun selbst schreibt
- Room 4
- Room 4
Plattformübergreifende Bereitstellung synthetischer Testdaten
In zunehmend hybriden IT-Landschaften stellt das Testen über Cloud- und On-Premises-Systeme hinweg eine Herausforderung dar – insbesondere dann, wenn echte Daten aus Datenschutz- oder Sicherheitsgründen nicht genutzt werden dürfen. Synthetische Testdaten bieten hier einen vielversprechenden Ausweg, doch häufig scheitern Initiativen an mangelnder Konsistenz und Kompatibilität zwischen verschiedenen Plattformen und Systemen.
Der Vortrag zeigt, wie ein zentral definiertes, fachlich orientiertes Datenmodell als Basis für die plattformübergreifende Generierung und Provisionierung synthetischer Testdaten in einer Privatbank dienen kann. Anhand konkreter Praxisbeispiele wird erläutert, wie testrelevante Daten über verschiedene Zielsysteme (z. B. Avaloq) hinweg konsistent generiert und flexibel bereitgestellt werden können – unabhängig davon, ob die Zielumgebung in der Cloud, On-Premises oder hybrid betrieben wird.
Durch die Anbindung verschiedener Testwerkzeuge und -frameworks (z. B. Tosca, UiPath, JUnit) ist gewährleistet, dass sämtliche Tools auf identische synthetische Testdaten zugreifen und diese flexibel bestellen können. Dies ermöglicht eine einheitliche und effiziente Testdatennutzung im gesamten Testprozess, unabhängig vom eingesetzten Automatisierungswerkzeug und unabhängig davon, ob in der Cloud oder On-Premise.
Teilnehmende erfahren,
wie sich ein zentrales, systemunabhängiges und fachlich geprägtes Datenmodell aufbauen lässt,
welche architektonischen Elemente eine flexible Provisionierung per API-, Pipeline- oder Self-Service-Mechanismen ermöglichen,
und wie sich Anforderungen an Datenschutz, Effizienz und Reproduzierbarkeit in Einklang bringen lassen.
Das Ergebnis: eine skalierbare, sichere und automatisierbare Bereitstellung synthetischer Testdaten für innovative, plattformübergreifende Teststrategien – mit Offenheit für die Integration verschiedenster Testwerkzeuge.
- Room 5
- Room 5
- Main Stage
How 1% Can Change Everything: Micro-Habits for Maintainable Test Automation
Small, consistent improvements beat big-bang changes. In this talk I show how the 1% principle makes test automation measurably more maintainable-step by step-and how AI acts as a co-pilot along the way. We’ll cover from stable selector design, architecture design and AI-assisted maintenance (e.g., test refactoring, flaky analysis, self-healing suggestions). At the same time, we’ll empower manual testers to safely use and contribute to automation with guardrails, clear prompts, and easy on-ramps in Playwright.
- Room 6
How 1% Can Change Everything: Micro-Habits for Maintainable Test Automation
Small, consistent improvements beat big-bang changes. In this talk I show how the 1% principle makes test automation measurably more maintainable-step by step-and how AI acts as a co-pilot along the way. We’ll cover from stable selector design, architecture design and AI-assisted maintenance (e.g., test refactoring, flaky analysis, self-healing suggestions). At the same time, we’ll empower manual testers to safely use and contribute to automation with guardrails, clear prompts, and easy on-ramps in Playwright.
- Room 6
11:15 - 11:30
Coffee Break
11:15 - 11:30
Coffee Break
11:30 - 12:00
11:30 - 12:00
When Testers Fix Bugs with Agentic AI
Today, testers report bugs — and wait for developers to fix them.
In our Lighthouse project, we’re exploring a future where testers no longer wait, but act: using agentic AI to analyze, generate, and apply bugfixes directly within the team’s development workflow.
This talk shares how we integrated tools like GitLab Duo, GitHub Copilot, and Telecontext (Deutsche Telekom’s MCP server) into our agile process so that testers can collaborate with AI instead of starting the traditional ticket cycle. You’ll learn how this changes roles, handovers, and team dynamics — and what happens when AI takes responsibility for real code changes.
At Swiss Testing Day, I will share openly whether this Lighthouse project was a success or not — including what worked well, what didn’t, and how we plan to adapt our approach in the future.
- Main Stage
- Main Stage
- Main Stage
E2E-Testing ohne Schmerz: Wie man Stabilität, Tempo und Qualität vereint
Die End-to-End-Testautomatisierung ist heute ein zentraler Bestandteil moderner Softwarequalitätssicherung. Doch gerade in umfangreichen Systemlandschaften stoßen viele Teams schnell an Grenzen: fragile Tests, komplexe Wartung und fehlende Skalierbarkeit führen oft dazu, dass Automatisierung ihr Potenzial nicht entfalten kann.
In diesem Vortrag zeige ich, wie durchdachte Architekturprinzipien, klare Teststrategien und gezielte organisatorische Maßnahmen zu einer nachhaltigen E2E-Testlandschaft führen. Aufbauend auf über zehn Jahren Praxiserfahrung in der Automatisierung von Unternehmensanwendungen werden praxiserprobte Methoden vorgestellt, mit denen sich Teststabilität, Wiederverwendbarkeit und Effizienz signifikant verbessern lassen.
Anhand realer Beispiele wird veranschaulicht, wie Teams ihre Tests so gestalten können, dass sie auch bei wachsender Komplexität wartbar und zuverlässig bleiben – unabhängig von Plattform, Browser oder Framework. Der Vortrag liefert konkrete Handlungsempfehlungen, mit denen Sie Ihre Testautomatisierung strategisch weiterentwickeln und langfristig erfolgreich betreiben können.
- Room 3
E2E-Testing ohne Schmerz: Wie man Stabilität, Tempo und Qualität vereint
Die End-to-End-Testautomatisierung ist heute ein zentraler Bestandteil moderner Softwarequalitätssicherung. Doch gerade in umfangreichen Systemlandschaften stoßen viele Teams schnell an Grenzen: fragile Tests, komplexe Wartung und fehlende Skalierbarkeit führen oft dazu, dass Automatisierung ihr Potenzial nicht entfalten kann.
In diesem Vortrag zeige ich, wie durchdachte Architekturprinzipien, klare Teststrategien und gezielte organisatorische Maßnahmen zu einer nachhaltigen E2E-Testlandschaft führen. Aufbauend auf über zehn Jahren Praxiserfahrung in der Automatisierung von Unternehmensanwendungen werden praxiserprobte Methoden vorgestellt, mit denen sich Teststabilität, Wiederverwendbarkeit und Effizienz signifikant verbessern lassen.
Anhand realer Beispiele wird veranschaulicht, wie Teams ihre Tests so gestalten können, dass sie auch bei wachsender Komplexität wartbar und zuverlässig bleiben – unabhängig von Plattform, Browser oder Framework. Der Vortrag liefert konkrete Handlungsempfehlungen, mit denen Sie Ihre Testautomatisierung strategisch weiterentwickeln und langfristig erfolgreich betreiben können.
- Room 3
- Main Stage
Why AI Is Useless for Compliance
…and Why You Still Need It
A strategic view on the current limitations of AI in legal compliance — and how to use it wisely.
While AI systems are rapidly transforming business operations, their application in legal compliance remains deeply constrained. From the inability to interpret legal nuance to the lack of transparency in decision-making, today’s AI is often misaligned with the core demands of regulatory work: accountability, context, and traceability.
This session offers a clear, expert-driven perspective on the state of AI in compliance — beyond the hype. Drawing on real-world cases and strategic frameworks, I outline the current technological boundaries, discuss key risks for organizations and regulators, and propose actionable approaches to using AI safely and effectively in complex legal environments.
Attendees will gain:
– A grounded understanding of what today’s AI can and cannot do in compliance
– Strategic considerations for implementing AI tools responsibly
– Case-based insights into where AI has failed — and where it adds value
– A framework for human-centered, auditable, and compliant AI integration
This talk is designed for: compliance leaders, general counsel, regulatory professionals, and executives seeking clarity — and confidence — in how AI can support (but not replace) human judgment in high-stakes domains.
- Room 4
Why AI Is Useless for Compliance
…and Why You Still Need It
A strategic view on the current limitations of AI in legal compliance — and how to use it wisely.
While AI systems are rapidly transforming business operations, their application in legal compliance remains deeply constrained. From the inability to interpret legal nuance to the lack of transparency in decision-making, today’s AI is often misaligned with the core demands of regulatory work: accountability, context, and traceability.
This session offers a clear, expert-driven perspective on the state of AI in compliance — beyond the hype. Drawing on real-world cases and strategic frameworks, I outline the current technological boundaries, discuss key risks for organizations and regulators, and propose actionable approaches to using AI safely and effectively in complex legal environments.
Attendees will gain:
– A grounded understanding of what today’s AI can and cannot do in compliance
– Strategic considerations for implementing AI tools responsibly
– Case-based insights into where AI has failed — and where it adds value
– A framework for human-centered, auditable, and compliant AI integration
This talk is designed for: compliance leaders, general counsel, regulatory professionals, and executives seeking clarity — and confidence — in how AI can support (but not replace) human judgment in high-stakes domains.
- Room 4
- Main Stage
Continuous Deployment in der Praxis
Continuous Deployment (CD) ist eine Softwareentwicklungs-Praxis, bei der Änderungen am Code automatisch in die Produktionsumgebung übertragen werden, sobald sie alle Tests der Delivery Pipeline (CI/CD) erfolgreich bestanden haben.
Dies ermöglicht häufige und zuverlässige Releases, reduziert manuelle Eingriffe und beschleunigt den Entwicklungsprozess.
Continuous Deployment (CD) erfordert eine grundlegende Veränderung in der Teststrategie, um die Effizienz und Qualität zu gewährleisten. Die Testausführungszeit muss drastisch reduziert werden, idealerweise auf maximal eine Stunde, um häufige Release-Zyklen zu ermöglichen.
In diesem Vortrag wird Torsten Welp erläutern, wie die Mobiliar diese Herausforderungen meistert und welche weiteren Hürden wir noch zu bewältigen haben. Diesen Vortrag habe ich an der Swiss Testing Night im Dez 2024 in Bern gehalten. Für mich ist das nach wie vor ein aktuelles Thema. Den Inhalt würde ich aktualisieren und verfeinern.
- Room 5
Continuous Deployment in der Praxis
Continuous Deployment (CD) ist eine Softwareentwicklungs-Praxis, bei der Änderungen am Code automatisch in die Produktionsumgebung übertragen werden, sobald sie alle Tests der Delivery Pipeline (CI/CD) erfolgreich bestanden haben.
Dies ermöglicht häufige und zuverlässige Releases, reduziert manuelle Eingriffe und beschleunigt den Entwicklungsprozess.
Continuous Deployment (CD) erfordert eine grundlegende Veränderung in der Teststrategie, um die Effizienz und Qualität zu gewährleisten. Die Testausführungszeit muss drastisch reduziert werden, idealerweise auf maximal eine Stunde, um häufige Release-Zyklen zu ermöglichen.
In diesem Vortrag wird Torsten Welp erläutern, wie die Mobiliar diese Herausforderungen meistert und welche weiteren Hürden wir noch zu bewältigen haben. Diesen Vortrag habe ich an der Swiss Testing Night im Dez 2024 in Bern gehalten. Für mich ist das nach wie vor ein aktuelles Thema. Den Inhalt würde ich aktualisieren und verfeinern.
- Room 5
- Main Stage
SiL-First-Ansatz im Testökosystem von AMG als Erfolgsfaktor
Die Inhouse-Softwareentwicklung spielt in der Automobilindustrie eine zunehmend zentrale Rolle, da innovative Softwarefeatures marktdifferenzierende Komponenten sind. Vor dem Einsatz im Fahrzeug durchlaufen sie unzählige Tests in diversen Umgebungen nach neuesten Methoden. Mercedes-AMG und tracetronic zeigen, wie für ein Steuergerät im e-Drive der SiL-First-Ansatz genutzt wird, um Fehler im Code frühestmöglich zu identifizieren und beheben zu können.
- Room 6
SiL-First-Ansatz im Testökosystem von AMG als Erfolgsfaktor
Die Inhouse-Softwareentwicklung spielt in der Automobilindustrie eine zunehmend zentrale Rolle, da innovative Softwarefeatures marktdifferenzierende Komponenten sind. Vor dem Einsatz im Fahrzeug durchlaufen sie unzählige Tests in diversen Umgebungen nach neuesten Methoden. Mercedes-AMG und tracetronic zeigen, wie für ein Steuergerät im e-Drive der SiL-First-Ansatz genutzt wird, um Fehler im Code frühestmöglich zu identifizieren und beheben zu können.
- Room 6
12:00 - 13:20
Lunch and Exhibition and Community
12:00 - 13:20
Lunch and Exhibition and Community
13:20 - 14:10
13:20 - 14:10
- Main Stage
Spring Cleaning for Test Managers: Lets Throw Out the Redundant Tests!
Large test suites often contain many tests that are absolutely redundant. They overlap with other tests so much that they only find bugs that these other tests already find. Such redundant tests are not only useless, but harmful: They slow down test execution without adding value in terms of bug-finding capability. And even worse, they eat up effort to maintain that would be better spent elsewhere. So let’s remove them!
But how can we detect redundant tests in our large test suites? And how can we be sure that they really provide no value?
In my talk, I will present several approaches to identify redundant tests. I will share their research background, our empirical studies and our experiences with how well they work under which conditions.
- Main Stage
Spring Cleaning for Test Managers: Lets Throw Out the Redundant Tests!
Large test suites often contain many tests that are absolutely redundant. They overlap with other tests so much that they only find bugs that these other tests already find. Such redundant tests are not only useless, but harmful: They slow down test execution without adding value in terms of bug-finding capability. And even worse, they eat up effort to maintain that would be better spent elsewhere. So let’s remove them!
But how can we detect redundant tests in our large test suites? And how can we be sure that they really provide no value?
In my talk, I will present several approaches to identify redundant tests. I will share their research background, our empirical studies and our experiences with how well they work under which conditions.
- Main Stage
Seven Trials: Rise of the Human Tester
Step into the fantasy kingdom of Softwarica, where AI promises perfection but hidden flaws threaten the realm.
In this Dungeons & Dragons–style simulation game, participants become heroes — Test Leads, Developers, Product Owners, Automation Specialists, and Stakeholders — navigating three thrilling Trials of Testing.
Through roleplay, dice-driven decisions, and collaborative problem-solving, participants will:
– Experience the tension between AI automation and human judgment
– Make strategic choices under limited resources and conflicting priorities
– Face ethical and human-centered challenges in software quality
The session turns abstract testing principles into a hands-on, engaging adventure: each decision affects outcomes, surprises emerge from dice rolls, and reflection is built into the gameplay.
By the end, participants will understand Human-in-the-Loop testing, appreciate the value of collaboration and oversight, and leave inspired to apply these principles in real-world QA environments — all while having fun in a fantasy adventure setting.
- Room 4
- Room 4
- Main Stage
Verborgene Angriffsfläche: Wie übersehene Schwachstellen die Schweiz Cyberangriffen aussetzen (25 min)
Für Banken und E-Commerce-Unternehmen ist es schon lange klar: E-Banking und Onlineshops müssen auf Schwachstellen untersucht werden – sonst wird es teuer. Doch wer testet Fotovoltaikanlagen, kabellose Tastaturen, weitverbreitete Apps wie TikTok oder Kernapplikationen in Schweizer Spitälern? Beispiele aus dem Arbeitsalltag des NTC zeigen, warum gewisse dringend notwendige Überprüfungen in der Schweiz nicht stattfinden und was dagegen unternommen wird.
- Room 6
Verborgene Angriffsfläche: Wie übersehene Schwachstellen die Schweiz Cyberangriffen aussetzen (25 min)
Für Banken und E-Commerce-Unternehmen ist es schon lange klar: E-Banking und Onlineshops müssen auf Schwachstellen untersucht werden – sonst wird es teuer. Doch wer testet Fotovoltaikanlagen, kabellose Tastaturen, weitverbreitete Apps wie TikTok oder Kernapplikationen in Schweizer Spitälern? Beispiele aus dem Arbeitsalltag des NTC zeigen, warum gewisse dringend notwendige Überprüfungen in der Schweiz nicht stattfinden und was dagegen unternommen wird.
- Room 6
- Main Stage
In the Agentic Era, QA Is Your Last Line of Defense. Insights from Banking, Insurance and Media Platforms (25 min)
AI agents are shipping software at unprecedented speed. Traditional test automation cannot keep up. In the Agentic Era, QA becomes the final control layer before production.
Built by the team that helped create Google Lens, our approach to agentic QA is fundamentally different. Instead of brittle scripts, autonomous agents understand products, reason about user flows, and continuously adapt to change.
Drawing from large-scale Banking, Insurance, and Digital Media platforms, this keynote shows how leading enterprises deploy best-in-class agentic QA to safeguard revenue, regulatory compliance, and customer trust. You will see what scales in agent-driven environments, and why autonomous QA is becoming core infrastructure across highly regulated industries.
In an agentic world, quality is no longer a testing problem. It is a strategic control system.
- Room 6
In the Agentic Era, QA Is Your Last Line of Defense. Insights from Banking, Insurance and Media Platforms (25 min)
AI agents are shipping software at unprecedented speed. Traditional test automation cannot keep up. In the Agentic Era, QA becomes the final control layer before production.
Built by the team that helped create Google Lens, our approach to agentic QA is fundamentally different. Instead of brittle scripts, autonomous agents understand products, reason about user flows, and continuously adapt to change.
Drawing from large-scale Banking, Insurance, and Digital Media platforms, this keynote shows how leading enterprises deploy best-in-class agentic QA to safeguard revenue, regulatory compliance, and customer trust. You will see what scales in agent-driven environments, and why autonomous QA is becoming core infrastructure across highly regulated industries.
In an agentic world, quality is no longer a testing problem. It is a strategic control system.
- Room 6
- Main Stage
Workshop: KI als Ansatz zur Reduktion von aufwändigen Barrierefreiheits-Tests
Der «European Accessibility Act» (EAA) ist am 28. Juni 2025 in Europa planmäßig in Kraft getreten. Er besagt, dass Software, elektronische Geräte und mobile Apps für Menschen mit Behinderungen, ältere Menschen oder Menschen mit geringer Technikaffinität zugänglich sein müssen – sei es im privaten oder öffentlichen Sektor. Bei der Schweiz kommt bereits das Diskrimierungsgesetz zum Tragen, und eine Gesetzesinitiative ähnliches des EAA ist bereits in Arbeit.
Dennoch legen viele Unternehmen nur einen geringen Wert auf dieses wichtige Thema, zumal die Strafen oft in Kauf genommen werden. Zumal aber eine bewusste Ausgrenzung von Menschen ethisch nicht vertretbar ist.
Ein Kostentreiber für Barrierefreiheit sind die umfangreichen manuellen Tests, die dafür benötigt werden, die Sicht des Endnutzers zu validieren. Bei den Kriterien zur Barrierefreiheit handelt es sich um nicht-fachliche Anforderungen, und diese sind nur schwer messbar.
Dieser Workshop zeigt, wie KI (Large Language Models) helfen kann, die Sichtweise von Endnutzern zu simulieren und manuelle Tests zu ergänzen, die sich auf den hier weltweit geltenden Standard beziehen (Web Content Accessibility Guidelines [WCAG]) und deren testbare Kriterien.
Der Workshop zeigt, was KI heute im Bereich Barrierefreiheit leistet, wie ein potenzielles Zielframework aussieht, und wie dedizierte Prüfungen zu den WCAG-Kriterien über ein LLM und Prompting durchgeführt werden.
- Cube 1
Workshop: KI als Ansatz zur Reduktion von aufwändigen Barrierefreiheits-Tests
Der «European Accessibility Act» (EAA) ist am 28. Juni 2025 in Europa planmäßig in Kraft getreten. Er besagt, dass Software, elektronische Geräte und mobile Apps für Menschen mit Behinderungen, ältere Menschen oder Menschen mit geringer Technikaffinität zugänglich sein müssen – sei es im privaten oder öffentlichen Sektor. Bei der Schweiz kommt bereits das Diskrimierungsgesetz zum Tragen, und eine Gesetzesinitiative ähnliches des EAA ist bereits in Arbeit.
Dennoch legen viele Unternehmen nur einen geringen Wert auf dieses wichtige Thema, zumal die Strafen oft in Kauf genommen werden. Zumal aber eine bewusste Ausgrenzung von Menschen ethisch nicht vertretbar ist.
Ein Kostentreiber für Barrierefreiheit sind die umfangreichen manuellen Tests, die dafür benötigt werden, die Sicht des Endnutzers zu validieren. Bei den Kriterien zur Barrierefreiheit handelt es sich um nicht-fachliche Anforderungen, und diese sind nur schwer messbar.
Dieser Workshop zeigt, wie KI (Large Language Models) helfen kann, die Sichtweise von Endnutzern zu simulieren und manuelle Tests zu ergänzen, die sich auf den hier weltweit geltenden Standard beziehen (Web Content Accessibility Guidelines [WCAG]) und deren testbare Kriterien.
Der Workshop zeigt, was KI heute im Bereich Barrierefreiheit leistet, wie ein potenzielles Zielframework aussieht, und wie dedizierte Prüfungen zu den WCAG-Kriterien über ein LLM und Prompting durchgeführt werden.
- Cube 1
- Main Stage
Workshop: Don’t trust a test you've never seen fail: Introducing Reverse Mutation Testing
Every day, companies rely on passing automated tests to release software updates quickly. But how can we trust these tests, if we’ve never seen them fail? Reverse Mutation Testing (RMT) helps validate your tests by deliberately breaking them to uncover false positives, catching faulty tests before they let real bugs slip into production.
As currently AI is generating code and even entire test suites, this question becomes even more critical: are AI-written tests truly effective, or just giving us a false sense of security? RMT provides the missing safeguard: ensuring both human and AI-generated tests can be trusted. With AI, we can go even further: intelligent mutation generation as well as AI-assisted detection of fragile test cases.
Learn how to apply this powerful approach and make your tests genuinely trustworthy. Because in the AI decade, testing your tests is not optional: it is essential.
- Cube 2
Workshop: Don’t trust a test you've never seen fail: Introducing Reverse Mutation Testing
Every day, companies rely on passing automated tests to release software updates quickly. But how can we trust these tests, if we’ve never seen them fail? Reverse Mutation Testing (RMT) helps validate your tests by deliberately breaking them to uncover false positives, catching faulty tests before they let real bugs slip into production.
As currently AI is generating code and even entire test suites, this question becomes even more critical: are AI-written tests truly effective, or just giving us a false sense of security? RMT provides the missing safeguard: ensuring both human and AI-generated tests can be trusted. With AI, we can go even further: intelligent mutation generation as well as AI-assisted detection of fragile test cases.
Learn how to apply this powerful approach and make your tests genuinely trustworthy. Because in the AI decade, testing your tests is not optional: it is essential.
- Cube 2
14:10 - 14:25
Coffee Break
14:10 - 14:25
Coffee Break
14:25 - 15:15
14:25 - 15:15
Programmer, entrepreneur, author, language designer, researcher, software legal expert
Panel Discussion: How Do We Define Quality in a Dangerous Decade? Becoming the Guardians of Trust in the Age of AI
As artificial intelligence becomes embedded in every layer of digital systems, quality assurance faces its most critical transformation yet. In this panel discussion, industry leaders explore how quality must evolve beyond functional correctness to address ethics, trust, and responsibility in an AI-driven world.
“How Do We Define Quality in a Dangerous Decade?” examines the balance between automation and human judgment, speed and accountability, innovation and oversight. The discussion challenges testers, developers, and decision-makers to step into a new role: Guardians of Trust.
Together, the panel will address how we test autonomous systems, govern AI decisions, detect bias and ethical risk, and ensure transparency and explainability across the software lifecycle. This session is a call to action for the QA community to safeguard not just software quality, but societal trust in technology.
Because in the age of AI, quality is no longer just a technical concern, it is a human responsibility.
- Main Stage
- Main Stage
Moderated by
Sponsored by
Testdaten und KI: Grenzen und Potenziale
Künstliche Intelligenz verändert die Welt des Testens weiterhin grundlegend – oft mit dem Versprechen, Komplexität drastisch zu reduzieren und Automatisierung zu vereinfachen. Doch wenn es um verlässliche Testdaten geht, stößt KI schnell an ihre Grenzen. Besonders das Reproduzieren seltener Fehler oder komplexer Ereigniskonstellationen erfordert die feine Detailtreue von Produktionsdaten – etwas, das KI alleine nicht leisten kann. Hinzu kommt eine häufig unterschätzte organisatorische Herausforderung: Unterschiedliche Datenkonsumenten wie Entwickler, Tester, Automatisierungsexperten oder ML-Ingenieure haben jeweils unterschiedliche Anforderungen. In diesem Spannungsfeld sind sowohl Fachwissen als auch Praxiserfahrung auf Seiten der Datenbereitstellung und -nutzung entscheidend. KI und Automatisierung können hier gezielt entlasten – vorausgesetzt, sie werden mit Bedacht und im Rahmen eines ganzheitlichen Ansatzes eingesetzt.
In diesem aufschlussreichen Vortrag beleuchten Christoph Stock und Jan Goebeler die Herausforderungen der Testdatenbereitstellung über verschiedene Umgebungen hinweg – abgestimmt auf die spezifischen Bedürfnisse der jeweiligen Konsumenten. Sie zeigen auf, wie sich technische und organisatorische Komplexität trennen lassen und in welchen Szenarien Künstliche Intelligenz echten Mehrwert bietet.
Zentrale Erkenntnisse:
Die Feinheiten der Testdaten-Orchestrierung: Lernen Sie die vielfältigen Herausforderungen kennen, Testdaten bereitzustellen – für verschiedene Nutzergruppen und unter Berücksichtigung von Qualität, Sicherheit, Verlässlichkeit und Relevanz.
Produktionsdaten vs. KI-generierte Daten: Erfahren Sie, wann KI-generierte Daten sinnvoll eingesetzt werden können – und wann reale Produktionsdaten unverzichtbar sind.
Ergänzen statt Ersetzen: Entdecken Sie, wie Sie KI gezielt einsetzen, um Ihre Testdatenlandschaft zu erweitern – z. B. durch intelligente Datenanreicherung oder Volumenerhöhung.
Zukunftssichere Integration Lernen Sie einen ausgewogenen Ansatz kennen, der die Innovationskraft KI-gestützter Lösungen mit der bewährten Verlässlichkeit von Produktionsdaten kombiniert – für Compliance, Performance und nachhaltige Qualität.
Nehmen Sie teil an einer praxisorientierten Diskussion, die den Hype hinter sich lässt und sich auf die realen Herausforderungen konzentriert. Ob Tester, QA-Experte, ML-Ingenieur oder Technologieführer – diese Session unterstützt Sie dabei, fundierte Entscheidungen zu treffen und von reaktiver Datenbereitstellung zu einer vorausschauenden Testdatenstrategie zu wechseln.
Nutzen Sie ein intelligentes, hybrides Modell: Lassen Sie KI die wertvollen Erkenntnisse aus Produktionsdaten ergänzen und steigern Sie mit Automatisierung Ihre Produktivität – für kürzere Durchlaufzeiten und eine bessere Developer Experience.
- Room 4
- Room 4
- Main Stage
Navigating the Maze: A Hunter's guide to successful Bug Bounty programs
The transition from classic security audits to Bug Bounty is often fraught with misplaced expectations and design pitfalls. This talk aims to navigate the practical realities of bug hunting, and how to turn a flood of reports (or lack thereof) into actionable intelligence. I will begin by defining the Bug Bounty ecosystem, identifying its key players and their motivations, while clarifying the fundamental differences between a bug bounty program and a traditional penetration test. I will share insights about what the life of a bug hunter really is, and the challenges to overcome to stay ahead in the race. We will then see common pitfalls and misconceptions that can arise when designing a bug bounty program, as they significantly differ from traditional quality testing. In particular, we will address the gap between a company’s perceived exposure and the intricate maze uncovered by hunters. I will provide a roadmap to launch a successful program that turns hackers into long-term partners, helps gain valuable insights for companies, and fosters security champions. We will conclude with a forward-looking perspective on the current state of bug bounty, and why Artificial Intelligence, while transformative, will not “solve” security but may instead introduce a new frontier of complex logic vulnerabilities.
- Room 5
Navigating the Maze: A Hunter's guide to successful Bug Bounty programs
The transition from classic security audits to Bug Bounty is often fraught with misplaced expectations and design pitfalls. This talk aims to navigate the practical realities of bug hunting, and how to turn a flood of reports (or lack thereof) into actionable intelligence. I will begin by defining the Bug Bounty ecosystem, identifying its key players and their motivations, while clarifying the fundamental differences between a bug bounty program and a traditional penetration test. I will share insights about what the life of a bug hunter really is, and the challenges to overcome to stay ahead in the race. We will then see common pitfalls and misconceptions that can arise when designing a bug bounty program, as they significantly differ from traditional quality testing. In particular, we will address the gap between a company’s perceived exposure and the intricate maze uncovered by hunters. I will provide a roadmap to launch a successful program that turns hackers into long-term partners, helps gain valuable insights for companies, and fosters security champions. We will conclude with a forward-looking perspective on the current state of bug bounty, and why Artificial Intelligence, while transformative, will not “solve” security but may instead introduce a new frontier of complex logic vulnerabilities.
- Room 5
- Main Stage
Game: Ups and Downs of GenAI
Are you wondering how GenAI can support your testing tasks? Maybe you’ve already experimented with ChatGPT, Copilot, or other large language models — and had mixed results. Are you curious why some use cases succeed while others fall flat, but not in the mood for another “This is how you should do it” lecture? Then, this game is made for you.
We’ll be playing a classic game you probably know: snakes and ladders. Every time a player climbs a ladder, we’ll explore a use case that shows the power of GenAI in testing. When a player slides down a snake, we’ll dig into the challenges and limitations of using GenAI effectively.
By the end of the game, there will be one winner on the board — and many winners in the room, because you’ll all leave with a clear, practical understanding of how to apply GenAI to your own testing tasks right away.
- Cube 1
Game: Ups and Downs of GenAI
Are you wondering how GenAI can support your testing tasks? Maybe you’ve already experimented with ChatGPT, Copilot, or other large language models — and had mixed results. Are you curious why some use cases succeed while others fall flat, but not in the mood for another “This is how you should do it” lecture? Then, this game is made for you.
We’ll be playing a classic game you probably know: snakes and ladders. Every time a player climbs a ladder, we’ll explore a use case that shows the power of GenAI in testing. When a player slides down a snake, we’ll dig into the challenges and limitations of using GenAI effectively.
By the end of the game, there will be one winner on the board — and many winners in the room, because you’ll all leave with a clear, practical understanding of how to apply GenAI to your own testing tasks right away.
- Cube 1
- Main Stage
Workshop: Machine Learning for Testers: Train, Evaluate, and Detect Bias in AI Models
This hands-on workshop is designed specifically for software testers interested in stepping into the world of Machine Learning (ML) with practical, easy-to-follow exercises using Python. Attendees will learn the fundamentals of ML from a QA perspective, including how to train a simple spam classifier using real-world data, understand what data ML models require, and most importantly, how to effectively test these models.
The session will also cover critical ethical aspects such as identifying and mitigating bias in datasets and model predictions. Participants will engage in building test cases for AI models, gaining skills essential for ensuring AI systems are fair, robust, and trustworthy. No prior experience in programming or data science is required—just curiosity and a tester’s mindset.
This workshop offers unique value as it bridges traditional testing with emerging AI technologies, a topic not widely covered in standard QA training, making it highly relevant for the growing AI testing community.
- Cube 2
Workshop: Machine Learning for Testers: Train, Evaluate, and Detect Bias in AI Models
This hands-on workshop is designed specifically for software testers interested in stepping into the world of Machine Learning (ML) with practical, easy-to-follow exercises using Python. Attendees will learn the fundamentals of ML from a QA perspective, including how to train a simple spam classifier using real-world data, understand what data ML models require, and most importantly, how to effectively test these models.
The session will also cover critical ethical aspects such as identifying and mitigating bias in datasets and model predictions. Participants will engage in building test cases for AI models, gaining skills essential for ensuring AI systems are fair, robust, and trustworthy. No prior experience in programming or data science is required—just curiosity and a tester’s mindset.
This workshop offers unique value as it bridges traditional testing with emerging AI technologies, a topic not widely covered in standard QA training, making it highly relevant for the growing AI testing community.
- Cube 2
15:15 - 15:30
Coffee Break
15:15 - 15:30
Coffee Break
15:30 - 16:00
15:30 - 16:00
- Main Stage
Testmanagement aus KI-Perspektive
Künstliche Intelligenz (KI) ist omnipräsent und verändert, wie wir Software testen. Sie erfordert neue Ansätze für das Testen von KI-Systemen – und zugleich die Anpassung bestehender Vorgehensweisen durch den gezielten Einsatz von KI selbst.
Welche Aspekte muss ich als Testmanager bei der Konzeptionierung von Tests neu berücksichtigen? Wie kann ich generative KI-Technologien sinnvoll nutzen, um den Testprozess effizienter zu gestalten? Dieser Vortrag zeigt auf Basis der KI-spezifischen Qualitätseigenschaften aus ISO/IEC 25059, welche neuen Herausforderungen beim Testen von KI-basierten Systemen entstehen, wie sich diese in der Praxis adressieren lassen und in welchen Bereichen KI-gestützte Werkzeuge das tägliche Testen unterstützen können. Die Teilnehmer erhalten konkrete Impulse, wie sie ihre Teststrategien an die Ära der KI anpassen können.
- Main Stage
Testmanagement aus KI-Perspektive
Künstliche Intelligenz (KI) ist omnipräsent und verändert, wie wir Software testen. Sie erfordert neue Ansätze für das Testen von KI-Systemen – und zugleich die Anpassung bestehender Vorgehensweisen durch den gezielten Einsatz von KI selbst.
Welche Aspekte muss ich als Testmanager bei der Konzeptionierung von Tests neu berücksichtigen? Wie kann ich generative KI-Technologien sinnvoll nutzen, um den Testprozess effizienter zu gestalten? Dieser Vortrag zeigt auf Basis der KI-spezifischen Qualitätseigenschaften aus ISO/IEC 25059, welche neuen Herausforderungen beim Testen von KI-basierten Systemen entstehen, wie sich diese in der Praxis adressieren lassen und in welchen Bereichen KI-gestützte Werkzeuge das tägliche Testen unterstützen können. Die Teilnehmer erhalten konkrete Impulse, wie sie ihre Teststrategien an die Ära der KI anpassen können.
- Main Stage
From Business Relevance to Test Data Intelligence – AI as a Strategic Enabler
In complex domains like banking and insurance, where long-term customer trust and relationships are essential, a key challenge lies in designing business-relevant use cases and test data that reflect the real complexity of customer interactions. Personas serve as focal points for business goals — representing both typical and aspirational customer types across products, channels, and services.
Yet, the link between these business expectations and the technical implementation at data and system level is often missing. By combining data modeling with AI, this gap can be closed: AI enables us to describe personas in natural language and translate it into structured, synthetic test data – including e.g. plausible income, transactions, or credit profiles.
This bridges the gap between business logic and data reality, enabling automated, compliant and meaningful test scenarios. A session for everyone who wants to make testing not just faster, but smarter.
- Room 3
- Room 3
- Main Stage
Who Tests the Machines That Build the Machines?
As we begin to form virtual engineering teams — teams of specialized agents that plan, code, review, test and deploy — the very nature of software creation is changing. But when machines collaborate with other machines, who ensures that quality, safety, and intent are preserved?
In this provocative session, we explore how Quality Assurance must evolve to meet the era of AI-engineered applications. Drawing on experiences from IT-HUSET, we’ll demonstrate how AI-driven QA frameworks can act as a wrapper around these virtual agent teams, automatically testing, verifying, and aligning their output with human-defined goals — all with AI as the enabler.
Yet, amid the automation, human testers remain essential. They bring judgment, context, and ethical oversight — the “human-in-the-loop” ensuring that our intelligent systems don’t just work, but work as intended.
Join us to rethink what quality means when your “team” isn’t entirely human. You’ll leave inspired — and perhaps slightly unsettled — about how we build trust in an era where AI engineers alongside us.
- Room 4
Who Tests the Machines That Build the Machines?
As we begin to form virtual engineering teams — teams of specialized agents that plan, code, review, test and deploy — the very nature of software creation is changing. But when machines collaborate with other machines, who ensures that quality, safety, and intent are preserved?
In this provocative session, we explore how Quality Assurance must evolve to meet the era of AI-engineered applications. Drawing on experiences from IT-HUSET, we’ll demonstrate how AI-driven QA frameworks can act as a wrapper around these virtual agent teams, automatically testing, verifying, and aligning their output with human-defined goals — all with AI as the enabler.
Yet, amid the automation, human testers remain essential. They bring judgment, context, and ethical oversight — the “human-in-the-loop” ensuring that our intelligent systems don’t just work, but work as intended.
Join us to rethink what quality means when your “team” isn’t entirely human. You’ll leave inspired — and perhaps slightly unsettled — about how we build trust in an era where AI engineers alongside us.
- Room 4
Finding Balance Between Coverage and Efficiency
Testing can feel like walking a tightrope between the light and the dark sides — between complete visibility and raw speed. On one side lies exhaustive coverage and perfect control; on the other, ruthless efficiency and delivery pressure.
The Light Side: complete visibility (high coverage, low efficiency) The Dark Side: raw speed (high efficiency, low coverage)
In this talk, we’ll explore how to find balance between these forces in modern testing. Drawing parallels from the Jedi philosophy, we’ll see how uncertainty isn’t an enemy to be eliminated, but a truth to be mastered.
Through real examples from SAP testing and automation at scale, we’ll discuss practical ways to balance coverage, cost, and confidence — and why chasing perfection may lead you to the dark side.
You’ll leave with new ways to visualize trade-offs, communicate uncertainty, and design testing approaches that flow — not fight — with reality.
“Perfect testing is like perfect knowledge of the Force — theoretically elegant, practically impossible.”
- Room 5
- Room 3
- Main Stage
QA is not enough anymore in the age of AI: protect your apps!
Today’s chat bots make it easy to attack your mobile or web app. You can test and check all you want – no code is perfect. AIs are ruthless and WILL find issues, putting your customers data and your reputation at risk.
One solution is to protect and harden your app – which comes with its own set of challenges. Ever executed an automated test case against a protected app? Good luck!
We will address all of those challenges and discuss potential solutions, such as:
– let AI help with testing
– bring security & testing much closer together
– allow testing of protected apps
– automation automation automation
- Room 6
QA is not enough anymore in the age of AI: protect your apps!
Today’s chat bots make it easy to attack your mobile or web app. You can test and check all you want – no code is perfect. AIs are ruthless and WILL find issues, putting your customers data and your reputation at risk.
One solution is to protect and harden your app – which comes with its own set of challenges. Ever executed an automated test case against a protected app? Good luck!
We will address all of those challenges and discuss potential solutions, such as:
– let AI help with testing
– bring security & testing much closer together
– allow testing of protected apps
– automation automation automation
- Room 6
16:00 - 16:15
Coffee Break
16:00 - 16:15
Coffee Break
16:15 - 17:10
16:15 - 17:10
- Main Stage
Closing Keynote
Test Transcendence and the AI Apocalypse
Test Transcendence and the AI Apocalypse delivers a bold and inspiring message for software testers ready to elevate their impact. While AI serves as a compelling illustration of today’s transformational forces, this session is ultimately about something much bigger: how testers can apply their unique skills to model risk, surface high-value insights, and influence critical decisions.
Drawing from decades of real-world experience, Rob Sabourin challenges outdated perceptions of testers as defect specialists or delivery blockers. Instead, he positions testers as extraordinary communicators of value, advocates for customers and stakeholders, and trusted partners in navigating uncertainty. This session empowers testers to step confidently into conversations that shape outcomes, protect trust, and help organizations thrive amid constant change.
- Main Stage
17:10 - 17:30
17:10 - 17:30
- Main Stage