Jump to:

Keynote

Limitless Within Our Boundaries

Speakers: João Proença

Overview: “We could accomplish so much more if we had that…”
Sounds familiar?

In this fast changing world of building software, it’s easy to think we can only achieve our goals if we have access to a whole lot of things – that shiny new tool, those cutting edge skills, a group of people with that expertise…

However, sometimes we are so blinded by our lack of options that we fail to see the downsides of having too many of them. It’s called the paradox of choice.

I’ve seen this paradox happening around me my whole life – not just in tech, but in so many other areas: as a musician, as a film addict, in the supermarket… You name it!

This talk is not just about the paradox of choice, but also about how sometimes within the right boundaries is where our most interesting opportunities arise. In fact, magical stuff has been made throughout history when few options were available. Let me share some of those stories that influenced me as a professional, as well as some others of my own! I will show you what it means to be limitless within boundaries and hopefully inspire you to be the driver of the change you want to see!

Bio: João Proença comes from Lisbon, Portugal, and is a Quality Engineer at Ada Health. He has assumed various roles throughout his career in the past 18 years, including quality assurance, development, customer support and marketing. Finding innovative solutions for difficult problems is what drives him the most, so he is always eager to talk about how professionals are overcoming testing challenges around the world. Outside of IT, João is passionate about songwriting, movies and football.

Jump to top

AI Testing Track

AI Déjà Vu: Lessons from Automation

Speakers: Kayla Gillman

Overview: Across the globe, organizations are rapidly investing in AI to improve quality, speed delivery, and reduce cost – yet many are encountering a familiar problem: high expectations, slow progress, and limited business value. A decade ago, QA leaders faced this same challenge during the rise of test automation. In this session, Kayla Gillman, Global Head of Pre-Sales at TTC Global, draws on her experience leading enterprise automation and quality engineering initiatives across multiple regions to explore what AI adoption can learn from automation’s first decade. Using real-world examples and industry research, Kayla examines why both automation and AI initiatives stall – not because of tooling, but due to gaps in readiness, governance, skills, and alignment with business goals. Delegates will gain a practical framework for assessing maturity and learn how to embed AI responsibly into modern quality practices without sacrificing trust or human judgment. Attendees will leave with clear, actionable guidance for scaling automation and AI sustainably across complex, global delivery environments.

Bio: Kayla Gillman is Global Head of Pre-Sales and Senior Manager, Canada at TTC Global, where she leads enterprise quality engineering strategy across international transformation programs. Based in Canada and working closely with clients across Europe and North America, Kayla specializes in test automation, AI-assisted testing, and large-scale delivery enablement. She brings close to two decades of experience helping organizations evolve from manual testing foundations to continuous and intelligent automation at scale. Known for her pragmatic leadership style, Kayla focuses on aligning people, process, and technology to deliver measurable business value. She is a frequent speaker on automation maturity, AI readiness, and the evolving role of quality engineering in global digital transformation.

Moving Beyond Pixel Diffs with AI-Powered Visual Reasoning

Speakers: Risko Ruus

Overview: As AI systems move from experimentation to production, traditional testing approaches are no longer sufficient. Deterministic test cases struggle to capture non-deterministic behavior, and failures are often subtle—ranging from hallucinations to inconsistent outputs across similar inputs.
This talk introduces a practical framework for Human-in-the-Loop (HITL) evaluation, where human judgment is systematically integrated into the evaluation process to improve reliability, trust, and decision-making.
Drawing from real-world experience building AI evaluation systems across both online (production) and offline (internal) environments, we will explore how to design scalable HITL workflows, define meaningful evaluation metrics, and evaluate and help shape guardrails that balance automation with human oversight.
Attendees will learn how to move beyond pass/fail testing and adopt a layered approach combining automated checks, human evaluation, and continuous feedback loops to ensure AI systems behave as expected in real-world scenarios.

Bio: I am a software quality enthusiast with over 20 years of experience in various companies and software projects.

I enjoy both developing software and testing it (including test automation).

Example applications I have worked on include Nokia smartphones, Skype, and mobile betting apps. My goal is to remain cross-functional and continually learn various aspects of the software engineering craft.

Some of my recent passions include:

learning how to optimize web and mobile app performance and developing tools for detecting web performance regressions
developing AI-supported Visual Reasoning tools for automatically finding issues and bugs from images

My hobbies include learning Japanese, backcountry hiking, and running.

Beyond Test Cases: Building Reliable AI Systems with Human-in-the-Loop Evaluation

Speakers: Pranav Pandit

Overview: Visual testing is supposed to protect QA teams from the familiar “it looks wrong” bug, yet traditional pixel-diff approaches only show that something changed, not whether that change actually matters. As modern interfaces grow more dynamic and design systems become more complex, teams need smarter ways to detect meaningful visual regressions.

This talk presents a practical approach to automated visual bug detection using multimodal LLMs. Drawing on a real-world implementation, it shows how AI models from providers such as OpenAI, Anthropic, and Google can be orchestrated to analyze screenshots and identify issues that pixel-based tools often cannot interpret on their own. These include layout breaks, missing elements, accessibility concerns, color contrast problems, and platform-specific guideline violations.

The session explores how AI-driven visual analysis can move beyond pixel-perfect comparison toward semantic understanding, helping teams distinguish intentional UI changes from genuine defects. It also addresses one of the biggest challenges in visual testing at scale: false positives, demonstrating how agent-based review systems can reduce noise while still surfacing critical issues.

Attendees will leave with practical ideas for using multimodal AI to strengthen visual testing workflows and make automated UI validation more accurate, scalable, and useful.

Key Takeaways:
How to evolve from “pixel diffs” to impact-based automated visual feedback
Patterns that turn image feedback into structured results (what changed, where, severity, why it matters)
Tips for integrating automated LLM-powered visual feedback into existing automated UI test frameworks

Bio:I’m a Staff Quality Engineer at Robinhood Markets with over 15 years of experience leading quality engineering transformations across fintech, ride share, and cloud security domains. My work focuses on building scalable quality systems and advancing the role of Quality Engineering in the age of AI.
At Robinhood, I’ve designed and led Human-in-the-Loop (HITL) evaluation systems for AI-powered features in production, uncovering critical quality gaps including ~15% accuracy variance and measurable hallucination rates—and shaping the guardrails required for safe deployment. I’ve also driven large-scale improvements in mobile test stability across complex iOS and Android automation ecosystems.
I specialize in bringing efficiency and intelligence into QA workflows through AI, combining automated systems with human judgment to create reliable, scalable evaluation frameworks for modern applications.
Beyond my industry work, I serve as the San Francisco Chapter Lead for the Ministry of Testing, where I contribute to the global testing community through speaking, workshops, and thought leadership on AI-driven quality engineering.

When AI Thinks Fast and We Think Less: Why Testers Must Think Better

Speakers: Nithin SS

Overview: As AI reshapes software development, the “commodity” of testing, writing scripts, and checking boxes is being outsourced to machines. But AI is a “magic box” that generates probabilistic guesses, not certainties. In this evolving era, the tester’s true value isn’t their ability to test, but their ability to think about testing. This talk provides a metacognitive framework to help you spot gaps in both human and artificial reasoning, ensuring that while AI accelerates our work, our critical thinking remains the ultimate safeguard for quality.

Abstract:
As humans, thinking is our nature, but left to its own devices, it is often biased, distorted, and uninformed. In software testing, this is a liability; in the age of AI, it is a crisis.

While AI can generate code and test cases in seconds, it lacks the ability to doubt, to understand context, or to reason ethically. The risk for modern testers is “cognitive atrophy”, a reliance on convincing but potentially shallow AI outputs that lead to “convincing falsehoods.” The quality of our testing mirrors the quality of our thought: better thinking → better test design, bug detection, and stakeholder persuasion.

This talk combines cognitive science (two-system thinking and common heuristics), practical metacognitive techniques, and an engineer-friendly framework I’ve developed from real testing experience. You’ll get hands-on strategies to audit your own thinking, validate AI outputs, and design tests that catch both software defects and the reasoning errors that hide them. You will learn how to use AI as an assistant to augment your creativity without surrendering your judgment, providing a clear structure for problem-solving in a world where the most important tool is no longer the software you use, but the mind that directs it.

Bio: Nithin brings over a decade of IT experience, specialising in fostering high-performing teams and driving Quality Engineering excellence. He currently serves as the Director of QA at Lodgify. Beyond his corporate role, Nithin is the founder of Synapse QA, a community dedicated to making impactful quality learning accessible, affordable, and available. He is also a career and leadership coach, guiding tech professionals in their career development.

Exploratory Testing is still human.

Speakers: Blay Smauel

Overview: A team runs an AI testing tool against their application for an afternoon. It generates thousands of interactions, flags defects, and produces a clean report showing 87% coverage. The team celebrates and ships.

Within hours, a critical defect slips into production.

Not because the AI missed something in the code, but because the problem was never in the code to begin with. It existed in the gap between what the system did and what users expected it to do.

In this session, we explore that gap.

You will see the same feature tested twice: once by an AI-driven approach and once through human exploratory testing. The comparison reveals a clear difference, not just in what is found but in how each approach understands the system. AI explores behavior. Humans question intent, context, and consequence.

Many of the most impactful defects are not requirement failures. They are failures of timing, recovery, and user expectation. These are the areas where exploratory testing has always been strongest and where human judgment remains critical.

This presentation breaks down where AI testing adds value, where it falls short, and how teams can combine both approaches effectively. Rather than replacing exploratory testing, AI changes how we apply it.

You will leave with a practical way to decide when to rely on automation, when to explore, and how to use both without creating blind spots in your testing strategy. Because tools will keep changing. The thinking is still ours.

Bio: Blay Samuel is a senior quality engineer focused on understanding how software behaves in the real world, not just in controlled test conditions.

With a background spanning front-end development, product management, and product design, he brings a practical perspective to testing that goes beyond execution to thinking. His work centers on exploratory testing, risk-based decision-making, and helping teams understand where automation and AI add value and where human judgment remains essential.

Based in London, UK, he leads quality engineering within a multinational healthcare and risk management company, where his work directly impacts millions of users across multiple regions.

He believes that while tools will continue to change, the thinking behind testing is what ultimately determines quality.

Test Strategy Track

Fantastic Biases & Where to Find Them in Software Development

Speakers: João Proença & Michael Kutz

Overview: Why did all our test cases fail because of this simple bug? Nobody tried that out before? How did five people agree to implement this terrible feature? Why are our estimates always so far off? There are many possible answers to those questions and none of them will be the whole truth. However, certain common cognitive biases might play a main role in all of the events leading to those questions. We all have them. They help us to think faster, but they also make us less rational than we think we are. They hinder our best judgement! In this talk I’ll demonstrate some of the most severe biases, explain their background, point out how they typically influence our professional decisions, and suggest some strategies to mitigate their effect. Being able to recognize and overcome biases in us and others is a long, challenging road for anyone – you won’t be able to do that journey with this talk alone, but you’ll certainly take your first step!.

Bio: Michael Kutz is a Staff Engineer at Entropy Data and the creator of the open-source testing tool ApproveJ (approvej.org). Previously, he spent almost 12 years as a Principal Engineer at REWE digital, where he spearheaded test automation, SRE practices, and enterprise-wide QA standards. As an international speaker, Michael explores the fascinating intersection of code and psychology. In his talk “Fantastic Biases and Where to Find Them,” he reveals how cognitive biases subconsciously sabotage our sprint planning, testing, and architecture—and how teams can build countermeasures against these mental bugs. When he is not analyzing systems, he enjoys running, playing video games, and exploring sci-fi universes.

From Tester to Quality Platform Engineer

Speakers: Jen Cook

Overview: When my team faced growing complexity across services, traditional QA approaches such as centralized testing, manual test ownership, and isolated automation couldn’t scale. In this talk, I’ll share the story of how we transformed our approach using platform thinking. Through this real-world example, we’ll break down the process of building and implementing an embedded, extensible test framework. I’ll show how treating frameworks, data generation, and environment setup as internal products, not one-off projects, changed the way our team approached quality. Attendees will see how this shift enabled product teams to own their quality, improved adoption of automation, and created a foundation for scalable, proactive validation. I’ll highlight the lessons learned and explore what it means when Quality Engineers step into platform engineering roles. This includes new success metrics, new collaborations with development teams, and a fundamentally different quality model. By the end of the session, participants will have a concrete, experience-driven framework for evolving quality practices from reactive testing to scalable, embedded enablement.

Bio: Jen Cook is a software testing professional with over two decades of experience spanning QA management and automation engineering. She believes in sustainable testing practices that empower teams rather than overwhelm them. Currently based in rural New Brunswick, Canada, Jen has spent her career championing sustainable testing practices that make life easier for teams and deliver better software. Jen has a passion for bridging the gap between manual and automated testing, ensuring that testing processes are both efficient and effective. In addition to working in tech, she runs a small-town record store with her partner. In her spare time, you’ll likely find her crafting, gardening, or building small apps.

Your tests are lying to you!

Speakers: Massimo Forno

Overview: Modern distributed systems are extremely dynamic, interconnected, and constantly evolving.
Yet most testing strategies remain static.

Teams run thousands of tests. Pipelines stay green. And still, critical failures reach production.

This is not a tooling problem.
It’s a model problem.

Traditional testing measures coverage of predefined scenarios, not actual runtime risk. As systems change, the relevance of those tests quickly decays.

In this talk, we introduce Software Signal Engineering (SSE), an approach that uses runtime signals, code changes, historical data, and machine learning to guide test selection.

Inspired by principles from quantum mechanics, such as superposition, uncertainty, and entanglement, SSE models software systems as probabilistic states rather than fixed conditions.
Instead of assuming certainty, it continuously updates risk based on observed behavior.

The result is a shift from deterministic validation to probabilistic confidence:

fewer but more meaningful tests
faster feedback cycles
higher resilience in production

Because in modern systems, certainty is often just an illusion.

Bio: Massimo is a software engineer with over 25 years of experience in large-scale distributed systems across banking, fintech, sport and media.

Throughout his career, he kept seeing the same pattern: test suites pass, yet systems fail in production.

Instead of adding more tests, or refining them forever, he started questioning the model itself.

Today, his work explores a different approach, using runtime signals, observability, and probabilistic thinking to decide what really matters.

Because in modern systems, certainty doesn’t scale.

Bug Hunting in Plain Sight: Finding Risk in Everyday Artifacts

Speakers: Karen N Johnson

Overview: Testers look for bugs in the product; experienced testers know to look for clues around it.

In this talk, I’ll share how everyday artifacts—release notes, Jira tickets, and even Slack conversations—can become powerful sources for Exploratory Testing.

These artifacts often contain signals about recent changes, hidden assumptions, and areas of fragility that aren’t obvious through traditional test planning. They reflect how the system is evolving, where communication breaks down, and where risk quietly accumulates.

Through real-world examples, including a case where release notes revealed bug fixes unknown to the testing team, I’ll show how I use these signals to generate bug hunting ideas and guide Exploratory Testing in more focused, intentional ways.

These moments are not just surprises—they are opportunities for targeted investigation. They point to places where understanding is incomplete, where coordination has slipped, or where changes may have unintended effects.

This session will introduce practical ways to:

Use release notes as a starting point for bug hunting
Mine bug tracking systems for historical patterns that inform new test ideas
Identify process signals that indicate elevated risk
Turn everyday project artifacts into actionable inputs for exploratory testing

Attendees will leave with a repeatable approach for expanding their test ideas beyond the product itself—learning to see risk in plain sight.

Bio: Karen N. Johnson is an exploratory tester who specializes in finding risk beyond the obvious. Her work centers on using real-world signals—release notes, Jira tickets, defect patterns, and team communication—to guide investigation and uncover hidden problems.
She writes The Tester’s Notebook, a weekly Substack newsletter focused on the practice of testing, where she explores how testers think, observe, and make decisions in complex systems

A Test Engineer’s Guide to Solving the Inverted Test Pyramid

Speakers: Emily O’Connor

Overview: Most testing professionals have heard of the automation pyramid, yet a variety of organisations I’ve encountered find themselves in a position where they have too many end-to-end and UI tests. This problem isn’t new, but some of the tools (AI) and technologies (frameworks) we have available, are.

UI and end-to-end tests are slow (taking mins to run rather than seconds), flakey (causing hours of investigation, watching back Cypress replays and re-writing when selectors change) and often owned by a small number of testing professionals. So working towards the traditional test pyramid embodies shift-left testing by moving validation as early as possible in the SDLC. Rather than finding bugs in production (shift-right) or even in a test environment, prioritising the automation pyramid enables a way of working which catches bugs during development where they’re cheapest to fix.

This talk covers how I’ve worked to fix inverted test pyramids and the effect it really has on achieving org-wide quality. Because the answer to a UI or end-to-end heavy number of test cases isn’t another framework; it’s a deep understanding of the critical assertions that should be made about your system at every level of the test pyramid and adopting AI in a way that enables surfacing information about bugs before they impact our users.

Bio: Emily O’Connor is a Principal Quality Engineer with a decade of experience in tech. An automation-savvy tester with a sixth sense for bugs, Emily is motivated to help software engineering teams build the right thing and build it once, by understanding users technology needs and creating environments that facilitate testing for them enabled by CI/CD powered by atomic automated tests that highlight regression issues product managers care about fixing.