UXLx 2025 — Wrap Up

From 20 to 23 May, UXers from 33 countries gathered in beautiful Lisbon for four immersive days of learning, sharing, and reigniting creativity. UXLx 2025 brought together a global community eager to explore the latest in User Experience through 14 hands-on workshops and 10 inspiring talks.

Whether you joined us this year, couldn’t make it, or are curious about what you missed, this recap has you covered.

🤝 UXLx thrives on collaboration, and this article is no different — jump in, highlight your favourite moments, and share your key takeaways!

💡 Read through the article or click a workshop title to jump directly to the one you’re interested in.

20 MAY · FULL-DAY WORKSHOPS

• Advanced Design Systems, Brad Frost
• Human Centered AI Product Sprint, Sarah Tan

21 MAY · HALF-DAY WORKSHOPS

🕘 Morning:

• Designing for Uncertainty with AI by Mike Oren
• Accelerating Insights with AI — A Hands-On Workshop for UX Researchers and Innovators, John Whalen
• Power of Storyboarding — How to Present Research Findings and Inspire Action, Laura Pledger

🕑 Afternoon:

• Product Voice, Torrey Podmajersky
• Design for Impact — Research Methods to Uncover Impactful Insights, Erin Weigel
• How to Maintain a DesignOps Roadmap, Peter Boersma

22 MAY · HALF-DAY WORKSHOPS

🕘 Morning:

• Be Less Wrong — Decision-Making for Building Great Products, Jeff Whitlock
• Designing AI-enabled Services, Yulya Besplemennova
• Bodystorming AI, Jane Park Storm with Daniela Jones

🕑 Afternoon:

• Designer 2.0 — Amplifying Design Work with AI, Matias Vaara
• Culturalisation and International UX, Chui Chui Tan
• Accessibility Operations, Devon Persing

20 MAY · FULL-DAY WORKSHOPS

Advanced Design Systems by Brad Frost

Overview of the “Advanced Design Systems” workshop room

Atomic Design creator, Brad Frost kicked-off this edition of UXLx with a workshop for experienced design system practitioners that know the basics inside and out, and were looking to dive into the details of design system architecture and process that truly make or break a design system effort.

Design systems and products have a mutually influential relationship — each shapes and evolves the other. However, in complex organisations, this interaction is far more intricate, requiring a layered and nuanced ecosystem to support a growing portfolio of digital products. That’s where The Design System Ecosystem comes in. It’s composed of five hierarchical layers, some are optional and don’t necessarily apply to every organisation:

• Core Design System Layer — Common UI building blocks for the entire organisation;
• Technology-Specific Layer (optional) — Versions of components adapted for specific frameworks or platforms;
• Recipe Layer (optional) — Sets of compound components used consistently in certain contexts;
• Smart Components Layer (optional) — Components with embedded logic to make developers use easier;
• Product Layer — The final websites and apps delivered to users.

Core Design System — It gathers reusable components, patterns, and principles that help the organisation officially define how interfaces are designed and built, and provides access to components and styles.

• Design Tokens — Basic design definitions like colours, typography, spacing, shadows, etc., that define the brand’s visual language;
• Icons — Can be managed separately or together with tokens;
• UI Components — The heart of the system — reusable components like alerts, modals, inputs, etc.;
• Reference Website — The “showcase” of the design system: part marketing, part documentation, part support. Brings together all assets (tokens, icons, components, Storybook, etc.) in one place. Can be built from scratch or with tools like Zeroheight, which integrates content from Figma and Storybook to give a full view to designers and developers.

Technology-Specific Implementation — This layer adapts the core design system to specific technologies, helping tech teams use the components natively in their environments.

• Framework Wrapper Layer — Involves embedding Web Components into framework-specific code. Useful for compatibility, gradual migration, and developer convenience. Includes: code repository, specific Storybook, and distribution package;
• Native Layer — More complex due to the diversity of technologies and OS standards. Requires platform-specific libraries and packages. Support tools are less mature than in the web ecosystem;
• Other Non-Web Implementations — Very diverse technologies. Require custom code and repositories to manage shared UI.

Recipe Layer (Optional but Highly Recommended) — Allows teams to create reusable compound components based on the core design system components.

• Figma Recipe Libraries — Like “visual cookbooks” created and maintained by product teams with specific compositions;
• Recipe Code Repositories — Store the composite components in code (Web Components, React, etc.), adapted to the team’s stack;
• Recipe Storybook — Documentation, testing, and visualisation of the compound components, aligned with Figma;
• Recipe Packages — Compiled and published for development teams to use in their products;
• Recipe Reference Website — Product-specific style guide for proper recipe usage with examples, best practices, guidelines and instructions.

What does it help with?

• Decentralises the creation of compound components;
• Increases system scale and flexibility;
• Gives teams autonomy while maintaining consistency with the core design system;
• Acts as an incubator for new components.

Smart Components Layer — This is a place where the components and recipes get wrapped in logic in order to provide downstream development teams with drop-in, ready-to-use functional components and services. Smart components that:

• Connect to APIs, databases, or internal services;
• Already include common business logic.

Examples:

• Forms with validation and submission (e.g. React Hook Form);
• Payment components connected to a gateway;
• Search with suggestions (typeahead) connected to a database;
• Tables with sorting, filters, and search;
• Integration with analytics tools;
• CMS-ready components;
• Dynamic product grids.

These smart components and common services handle recurring logic and functionality, allowing development teams to focus on more relevant tasks.

Product Layer — This is where all the pieces of the design system come together to create real applications and websites that users interact with daily. It’s here that the true success of the design system is measured, it must be the foundational infrastructure that supports real digital products.

• Product design files — The design system’s Foundations library, with the appropriate theme applied. The design system’s UI component library. Any relevant recipe libraries;
• Product codebase powered by framework — This is where apps built with frameworks like Next.js, Remix, or Express live. It’s at this level that product-specific business logic, routing, state management, and caching are implemented;
• Product codebase not powered by framework — Products not using modern JavaScript frameworks can still consume Web Components from the design system directly via CDN or NPM. This allows the design system to power any web-based product, regardless of tech stack, with updates managed by simply pulling the latest library version;
• Non-web product codebases — Native apps built in non-web environments (Swift, Kotlin, etc.) can’t use Web Components directly. These platforms require their own versions of the component library, tailored for mobile operating systems and technologies.

When you’re finished changing.

You’re finished.

Human Centered AI Product Sprint by Sarah Tan

Example of an AI Capability Card (Machine Learning)

Sarah brought us a deep dive into the Human-Centered AI methodology to transform our approach to AI development. It consists of a 5-step framework that bridges the gap between human needs and AI capabilities:

1. Define AI-Enabled capabilities — Identify business opportunities, user painpoints, and areas where AI can add value (creating an AI opportunity statement);
2. Align user needs to data inputs — Align business and user needs to achievable data needs and AI inputs (AI tech — User value prioritisation matrix);
3. Ideate AI-enabled opportunities — Brainstorm and generate ideas for new possibilities enabled by AI capabilities (AI capability cards);
4. Explain product features to users — Communicate what AI does to manage user’s expectations and build trust;
5. Impact of tech on society — Consider impact of AI solutions across different aspects of society and evaluate unintended consequences (using AI ethics cards).

Welcome Party @ Esplanando 🍻☀️🌳

A short riverside walk brought us to a laid-back terrace with stunning views of the Tagus, the Vasco da Gama Bridge, and Portugal’s tallest building —Vasco da Gama tower. After a full day of hands-on learning, there’s nothing like unwinding with a drink in hand and great company all around. 😎🍻

21 MAY · HALF-DAY WORKSHOPS

Designing for Uncertainty with AI by Mike Oren

Mike Oren presenting the workshop

Mike’s workshop drawn on the idea that basic AI knowledge is essential for designers so his intent was to demystify how AI works and help designers feel more confident in conversation with data scientists and machine learning engineers.

We started off by clarifying some statistics basics and two key measures: Correlation and Causation.

• Correlation is when two variables change together, but that doesn’t mean one causes the other. It’s measured with an R value (Pearson’s correlation coefficient), where values closer to 1 indicate a stronger relationship. A value beyond ±0.7 is typically considered statistically significant.
• Causation means one thing directly affects another. This is determined through controlled experiments that isolate variables and account for confounding factors. It’s commonly measured with a p-value, where values closer to 0 indicate stronger evidence of a causal relationship. A p-value below 0.05 is often used as a threshold for significance.

In AI design:

• Traditional machine learning models often rely on causal reasoning (like regressions), assuming there’s a theoretical or experimental basis for predictions.
• Large Language Models (LLMs), on the other hand, identify patterns and correlations in massive text datasets, but don’t truly understand or infer causation.
• While newer LLMs can explain their “reasoning,” their outputs are ultimately based on probabilities, not causal logic.

Understanding Probabilistic Design in AI — Probability, the measure of how likely something is to happen (from 0 for impossible to 1 for certain), plays a key role in AI. LLMs, for instance, generate outputs based on word probabilities, making their behaviour inherently uncertain.

What is Probabilistic Design? — It’s a design approach that embraces uncertainty by using probability to anticipate a range of outcomes. This helps designers build more adaptable, personalised, and resilient user experiences.

Why does it matter in AI Design?
By understanding probability, designers can:

• Create safer, more transparent, and explainable AI systems;
• Better manage unexpected or undesired outputs;
• Design engaging and personalised interactions;
• Weigh potential risks and rewards more effectively.

By demystifying AI, as designers we become empowered to shape the experience that delivers value and reduces misuse & abuse.

Designers must balance unpredictability with safety, align AI outputs with human values, and evaluate the success of probabilistic approaches — tasks that require both technical and ethical considerations.

Demystifying probability empowers designers to shape AI experiences that are thoughtful, responsible, and valuable.

Design Researchers and Data Scientists — Designers working with data scientists can improve experiences. Together they can help uncover the key contexts that help turn knowledge into insights. Designers can craft interfaces that encourages trust that drives people to act on the insights.

Designers partnered with data scientists leads to improved context for smarter, more adaptable interfaces.

Accelerating Insights with AI — A Hands-On Workshop for UX Researchers and Innovators by John Whalen

John Whalen discusses with attendees during a workshop exercise

Initially sceptical about AI and how it could fit in the research process, John Whalen conducted a comprehensive investigation that allowed him to be confident enough to start leveraging AI in his research practice. Now the typical research process that would typically last 35 days over 7 weeks is now conducted in less than 7 days. AI has transformed every stage of the customer research process, from planning to sharing results:

1. Prioritising research — AI helps test hypotheses and report on past research using synthetic users and automation;

2. Designing the study — Custom GPTs support study design and recruitment planning. Synthetic users help test and refine study designs;

3. Recruiting — AI tools streamline and scale participant recruitment;

4. Data collection & synthesis — AI moderates interviews, creates summaries, and generates highlight reels to accelerate analysis;

5. Results share out — GPTs and AI tools help report findings, interrogate data, and simulate user discussions through video.

With practical examples and exercises, John walked us through the new AI skills we need:

• Prompt engineering — How to create a prompt to help us create an interview guide;
• AI-moderated interviews —Exploring the ListenLabs tool;
• AI-powered analysis — Giving our favourite LLM a research brief (business goal, research goal, and stakeholder questions to answer), a research analysis prompt and the interview scripts and see how they analyse it;
• Synthetic users — Creating a synthetic user with a LLM by providing them with our research brief and synthetic user prompt.

Finally, John reflected on the future of UX research roles:

AI isn’t replacing UX researchers. It’s reshaping the role.

While AI tools are automating tasks that used to took days or weeks, like transcription or analyse massive volumes of research data, the UXR role is shifting away from hands-on execution toward strategic orchestration. The UX researcher will evolve from collector to orchestrator, from note-taker to strategist, and from moderator to analyst and sensemaker.

Power of Storyboarding — How to Present Research Findings and Inspire Action by Laura Pledger

Laura presenting the workshop

How can storyboards impact my UX research work? How can I introduce it in my organisation? These couple questions were certainly in everyone’s minds at the start of Lauren’s workshop.

That’s why Lauren started by presenting us the concept of storyboards — a visual tool that help to capture the attention of an audience, and because they resonate on a human level, they can even help change perspectives.

Storyboards are incredibly useful because they are visual tools that help bring the user’s story to life, making abstract ideas more concrete and engaging. They capture people’s attention and draw on the empathy of the audience, allowing viewers to resonate with the story on a personal level. By providing additional context, storyboards help clarify concepts and are suitable for all levels of understanding, from beginners to experts. Moreover, stories presented through visuals tend to be more memorable than standalone facts, making storyboards a powerful communication and design tool.

Storytelling plays a crucial role in influencing emotions and decision-making. The brain actively seeks input from the body and vice-versa when making decisions, and our emotions significantly impact how rational those decisions are. In fact, without emotional input, our capacity for rational thought becomes impaired. By presenting information through stories, you can tap into the audience’s emotions, making your message more persuasive and helping you build trust as a presenter.

After walking us through a couple of case studies to better understand the storyboard concept, Lauren covered the story arc technique and how to use it to structure compelling stories and what are the elements of a good story.

Then it was time to get hands-on and create our own storyboard focusing on 5 steps:

1. Review research;

2. Plan your story — Defining the lead character, their goal or problem to solve, their pitfalls or obstacles, and which other characters are involved;

3. Create plotline — Beginning, middle and end. Rising and falling action;

4. Write narrative — It should be in the 3rd person and kept short;

5. Add visuals — People and facial expressions, thought and speech bubbles, icons and characters.

How can storyboards be used throughout the UX process?

• Research and usability testing
• Enriching journey maps
• Ideation
• Prioritisation

Product Voice by Torrey Podmajersky

Torrey discussing an exercise with a work group

Having just delivered the 2nd edition of her book Strategic Writing for UX, a best-seller in the UX field, Torrey joined us at UXLx focused on the voice of a product.

As we entered the workshop and took our seats, each table/group had a card describing a brand (description, industry, users, current priority). Through a series of guided exercises — defining the product principles (your impressions, brand words, competitors, differentiators), drafting and using a voice chart to rewrite a brand experience — our ultimate goal would be designing the product’s voice.

What’s product content? Product content refers to the information and materials that help users interact with and understand a product. The goal of product content is to enable users to accomplish their goals throughout the product journey, from investigating and verifying to engaging, troubleshooting, and eventually becoming product champions. Good product content is systematic and works at all levels to support the user’s experience with the product.

The cycle of UX content in a product journey involves several stages:

• Investigating and Verifying — Users explore and gather information about the product to determine if it meets their needs. Content like ads, product pages, and reviews play a crucial role here;
• Committing and Setting Up — Once users decide to use the product, they commit to it by signing up or purchasing. Content such as onboarding guides and setup instructions help them get started;
• Using — Users actively engage with the product. Content like UI text, in-app guidance, and how-to articles support them in using the product effectively;
• Fixing — If users encounter issues, troubleshooting guides and support content help them resolve problems and continue using the product;
• Becoming Product Champions — Satisfied users may become advocates for the product, sharing their positive experiences with others. Content that encourages sharing and provides badges or rewards can support this stage.

There isn’t a strict “default” for the cycle, as it can vary depending on the product and its users. However, the key is to ensure that content is strategically designed to support users at each stage, addressing their needs and enhancing their overall experience with the product.

Define voice for a product — This refers to the personality and tone conveyed through the product’s content. A consistent voice helps establish a brand’s identity and makes the product more relatable to users. It should align with the brand’s values and resonate with the target audience. For example, a playful and friendly voice might be suitable for a casual app, while a more formal and professional tone might be appropriate for financial software. Voice guidelines often include aspects like expressiveness, playfulness, and clarity, ensuring that all content reflects the desired personality.

Key components of establishing a consistent and recognisable voice in UX content:

• Concepts — The core ideas and themes that the content communicates should align with the brand’s identity and user expectations. Concepts should be clear and relevant to the user’s needs and the product’s purpose;
• Vocabulary — The choice of words should reflect the brand’s voice and be appropriate for the target audience. Using familiar and straightforward vocabulary can enhance understanding and engagement;
• Verbosity — This refers to the length and complexity of the content. Striking the right balance is crucial — content should be concise and to the point, avoiding unnecessary jargon or overly complex language that might confuse users;
• Grammar — Proper grammar ensures clarity and professionalism in content. It helps convey messages accurately and maintains the credibility of the brand;
• Punctuation & Capitalisation — Consistent punctuation and capitalisation contribute to the readability and visual appeal of content. They help guide users through the text and emphasise important points, ensuring that the content is easy to scan and understand.

By carefully considering these elements, UX content can effectively communicate with users, enhancing both the usability and the overall user experience of the product.

Use voice for content design — A voice chart is a tool used to define and maintain a brand’s voice across various types of content. It outlines specific guidelines for how the brand should communicate, ensuring consistency and clarity.

One voice, many possibilities.

Design for Impact — Research Methods to Uncover Impactful Insights by Erin Weigel

Erin during the “Design for Impact” workshop

Erin started by clearly stating the learning goals for the workshop:

• How to approach research to create real impact
• How to communicate research to expand your impact

We covered 4 main topics:

• Impactful research approach — How do we get from insights to impact?; Hierarchy of evidence; Discover real problems that when solved are likely to impact important metrics;
• How to ask good research questions — Four useful types of questions (exploratory, descriptive, explanatory, relationship-based); Good research questions are clear, specific, and goal-oriented; Use the right question type for what you need to learn;
• How to pick the right research method — Your question influences your research method; Different research methods give you different types of data; Use the Landscape of User Research Methods to choose your method based on your question type;
• Communication to inspire action — What’s the best way to share what you learn?; Keep reports short and sweet; Always reference your business goals and objects in your research; Communicate your insights to drive actionable impact within your business.

How to Maintain a DesignOps Roadmap by Peter Boersma

Peter discussing with a group of attendees during the workshop

As a DesignOps consultant, Peter defines DesignOps as:

“We continuously and structurally improve the circumstances for designers and the design org, to increase the chances of good design happening.”

This workshop taught us how to make sure the right DesignOps initiatives are executed with the right people, to have the biggest possible impact.

Peter walked us through a few examples of DesignOps initiatives from some major companies:

• Miro’s design process— The design process follows the larger product development process; the goal is to align design work between product teams;
• Miro’s design project trackers — The goal is to support Design Managers in planning and tracking work;
• Design review guidelines —The goal is to support designers in reviews;
• ServiceNow design team review — The Design Experience Review Team (DERT) supports designers in concept phase and design systems usage;
• Design bug robot — The goal is to discover design bugs faster and improve recovery time;
• Thought Leadership Council —A team of 3 people with the goal of maintaining thought leadership status;
• Invited presentations — The goal is to build shared culture and terminology;
• IKEA’s design operations partner —The goal is to optimise implementation of DesignOps services for a team

For the next part, we covered what DesignOps Managers do with DesignOps initiatives, with practical exercises for each:

1. Inventory — Continuously, and from internal and external sources;

2. Prioritise — With stakeholders and Ops peers;

3. Distribute — Define scope and teams;

4. Coordinate — Ensure progress, report, and adjust scope;

5. Measure— Prove the impact of DesignOps and Design;

…and when there is time left: Execute and iterate.

You should have a DesignOps Roadmap to make sure that the right DesignOps initiatives are executed in the right order and with the right people, to have the biggest possible impact.

22 MAY · HALF-DAY WORKSHOPS

Be Less Wrong — Decision-Making for Building Great Products by Jeff Whitlock

Jeff presenting the workshop

Jeff’s workshop focused on decision-making strategies, emphasising the importance of understanding reality accurately to make good decisions. It introduced a four-step decision process and discussed the differences between deliberate and default decision-making.

Jeff stated that this workshop was not about AI, but it’s more important than ever in the time AI — and why is that? The ability to be “less wrong” and make failures productive is more important than ever. It’s crucial in today’s fast-paced and complex world to understand reality accurately. This workshop provided a structured framework and practical tips to improve decision-making, helping individuals and organisations navigate high-consequence decisions effectively.

What is a decision? — Anything that leads to action. There are two types of decision-making:

• Deliberate decision-making, which is slow, thoughtful, and intentional;
• Default decision-making, which is mindless, habitual, or driven by the status quo.

Default decision-making often results from avoidance or organisational dysfunction, where no one clearly owns the decision, leading to the status quo prevailing. And this decisions can be:

• High vs Low Consequence
• One-time vs Repeat Decisions

Productive failure — The goal of being “less wrong” is to make fewer mistakes and to ensure that when mistakes do occur, they are productive. This involves learning from failures so that the value of the lessons learned is greater than the cost of the failure.

Be less wrong by:

• Using a good decision-making process — Implement a structured approach to decision-making, which includes determining the type of decision, selecting an appropriate approach, making the decision, and comparing outcomes to document learnings;
• Writing down your principles — Develop and document principles based on experience to guide future decisions;
• Making fewer, higher-value decisions — Focus on making fewer decisions that have a higher impact, rather than trying to make every decision perfectly. This involves concentrating on the most important decisions and using principles and mental models to guide them.

Additionally, to decrease the cost of being wrong:

• Smaller bets — Make smaller, less risky decisions to minimise potential negative impacts;
• Faster feedback cycles — Implement faster feedback loops to quickly learn from decisions and make necessary adjustments.

Avoid the 7 deadly sins of decision-making:

1. Making a bad decision and not recognising it was a bad decision;
2. Taking too long to make a good-enough decision;
3. Deciding by being indecisive;
4. Failing to identify when a decision is a flatline decision;
5. Flipping AND and OR decision options;
6. Prematurely limiting option set (convergence before divergence);
7. Making an ambiguous decision;

A 4-step process for decision making:

1. Determine the type of decision:

• Assess whether the decision is high or low consequence;
• Determine if it is a one-time or repeat decision;
• Map the decision to the appropriate quadrant.

2. Select an approach:

• Define the question and determine the permanence and uncertainty;
• Define success criteria;
• List all options, including absurd ones.

3. Make the decision:

• Document the decision clearly;
• Include the rationale and key determinants;
• Write down predicted outcomes.

4. Compare outcomes and document learnings:

• Review decisions quarterly;
• Compare predictions to actual outcomes;
• Extract lessons and principles;
• Update the decision journal.

Designing AI-enabled Services by Yulya Besplemennova

Yulya discussing with attendees during a workshop exercise

With Julia (together with Serena Talento) we had the chance to try out the AI Opportunity Landscape Canvas — a really useful tool for identifying where AI can add value to services and processes. The structured approach helped us explore different kinds of AI opportunities, from automation to assistance and augmentation. We also got to experiment with various tech enablers and think about how they could shape more innovative and human-centered experiences.

We started by getting familiar with the overall framework and discussing how AI can be applied in different contexts. It was great to reflect not just on the possibilities, but also on the interaction models, design patterns, and ethical questions that come with AI.

Then we looked at a real-world service example, mapped it along the value chain, and brainstormed ideas for how AI could be integrated using combinations of tech enablers.

The hands-on group work was definitely a highlight — using printed canvases and cards really helped spark conversation and collaboration.

Bodystorming AI by Jane Park Storm with Daniela Jones

Jane (on the left) and Daniela (on the right) delivering the workshop

This workshop was really a quite unusual and thought-provoking way of thinking about AI. Instead of just talking about generative AI, we embodied it — literally. Using a method called bodystorming, we physically acted out what it might feel like to be an algorithm: unpredictable, creative, sometimes nonsensical, and often surprising.

The session started with a mix of design thinking and improv techniques — especially the “yes, and” principle — which really helped loosen things up and made it easier to explore the weirdness and wonder of AI without judgment. It was both fun and unsettling to put ourselves in the shoes of a technology that could eventually outpace us.

Designer 2.0 — Amplifying Design Work with AI by Matias Vaara

Matias during the “Designer 2.0” workshop

Matias’ workshop explored the latest advancements in artificial intelligence and their practical applications in design, ideation, research, and prototyping. It provided an in-depth look at how AI tools are reshaping workflows, increasing productivity, and enabling more strategic, creative, and ethical decision-making.

Most useful uses of AI

• Research assistance — AI can conduct deep research and provide comprehensive analysis quickly, allowing designers to uncover insights and trends efficiently. Tools like OpenAI Suite, Dovetail, and Perplexity are particularly useful for this purpose;
• Creative ideation — AI can significantly boost creative ideation, especially in the early phases, by generating a large volume of initial ideas;
• Prototyping and creation — AI tools assist in drafting user flows, creating mockups, and generating UI designs, providing a starting point for further refinement, accelerating the prototyping process;
• Content generation — AI excels at the generation of written content, from copywriting and documentation to meeting summaries and presentation materials. It can produce high-quality text for various applications, from marketing materials to user interfaces.

Biggest risks

• Quality and reliability — AI models, while powerful, are not always 100% accurate and can produce errors or “hallucinations”;
• Data privacy — The use of AI, especially in handling sensitive data, has high risks to privacy;
• Creative homogenisation — There is a concern that widespread use of AI for ideation and creation could lead to homogenised outputs, as AI models may generate similar ideas or designs based on their training data;
• Professional displacement — AI has the potential to disrupt job markets and economic structures, leading to concerns about job displacement and the need for workforce to learn new skills and adapt to new roles that leverage human creativity and strategic thinking alongside AI;
• Over-dependence risks — There is a risk of becoming overly reliant on AI, which could lead to a loss of human creativity and critical thinking skills.

Where is generative AI most useful in design?

• Research — Use as research assistant. Deep research capabilities (10–20 min comprehensive analysis). Tools: OpenAI Suite, Dovetail, Perplexity;
• Learning — Understanding unfamiliar concepts. Developing personalised learning plans. Providing technical support;
• Ideation — Can boost creative ideation by 40% (BCG/MIT/Harvard study). Best for early-phase ideation. Generating large volume of initial ideas;
• Strategy — Analysing market trends. Conducting deep research;
• First drafts — Creating initial content drafts. Generating user flows and mockups;
• Summaries and feedback — Synthesising research findings. Summarising data.

3 C´s of prompting

Context:

• Establish clear goals for what you want to achieve with AI;
• Provide relevant background information to guide the AI’s understanding;
• Incorporate user insights to tailor the AI’s responses to specific needs;
• Use a specific lens or perspective, such as viewing the problem as an accessibility researcher or marketing specialist, to refine the AI’s approach.

Constraints:

• Define length requirements for the AI’s responses to ensure they are concise and to the point;
• Specify the tone you want the AI to adopt, whether formal, casual, etc.;
• Apply strategic frameworks to guide the AI’s reasoning and ensure alignment with broader objectives.

Continuity:

• Engage in iterative conversations with the AI to refine and improve outputs;
• Maintain context throughout interactions to build on previous insights and avoid starting from scratch;
• Explore different branches of a topic to uncover new insights and perspectives.

Explore ideias visually

• Image generation — Generate realistic images using AI models like GPT-4. Create early presentation images, mood boards, and persona images;
• Service storyboards illustrations — Create visual storyboards by generating scenes one by one. Visualise service interactions and user journeys;
• Visual ideation of future visions — Envision future scenarios with AI-generated images. Useful for speculative design and strategic planning;
• AI-assisted mockups — Use tools like Stitch by Google or UIsart.io to generate UI designs from text prompts. Provide a visual starting point for further refinement;
• Iterative visual refinement: Use AI to iterate on visual designs with feedback and adjustments. Focus on criteria like color schemes or layout preferences.

Ethical considerations

• Bias and discrimination — AI models can amplify biases present in their training data, leading to discriminatory outcomes;
• Privacy loss — The use of AI poses risks to privacy, especially when handling sensitive data;
• Economic disruption — AI has the potential to disrupt job markets and economic structures, leading to concerns about job displacement;
• Environmental impact — The computational resources required for AI can have a significant environmental footprint;
• Authority issues — Delegating decision-making to AI systems raises questions about accountability and transparency.

Recommendations

• Stay informed and adapt — Continuously update your knowledge of AI advancements and tools to remain competitive and leverage new capabilities effectively;
• Embrace AI as a creative partner — Form a creative partnership with AI, using it to amplify your strengths in empathy, intuition, and strategic thinking while benefiting from AI’s computational power and knowledge;
• Focus on strategic impact — Use AI to focus on higher-level strategic tasks, allowing you to take on a wider range of roles and responsibilities while enhancing efficiency and productivity;
• Consider ethical implications — Be mindful of ethical considerations, such as bias, privacy, and over-dependence, and strive to use AI responsibly and transparently;
• Experiment with emerging tools — Explore and experiment with new AI tools and methodologies to find the best fit for your workflows and organisational needs, staying ahead of the curve in AI adoption.

Culturalisation and International UX by Chui Chui Tan

Chui Chui presenting at the workshop

With more than 17+ working with major brands and corporations across multiple markets, Chui Chui started off by giving some examples of brands like Uber, Walmart, Starbucks or Sephora that couldn’t make it when entering new global markets, simply because they failed to understand the markets they were entering.

To bridge the gap between local cultures and global business you must move beyond basic localisation (like simply translation) to designing experiences that align with diverse cultural contexts and needs.

Instead, business must apply Culturalisation across business functions. There are three levels of Culturalisation:

• Level One — Respect establishment;
• Level Two — Cultural expectation (name formatting in forms, payment methods, for example);
• Level Three — Experience enhancement.

Culturalisation can influence design, marketing, customer support, product development and overall strategy to strengthen your global presence.

To gather cultural insights and make informed decisions and strategies to grow in a global market we can use The Four-Bucket ‘Market Knowledge Clarity” exercise. This is a place where you put together known facts (high certainty), strong hypotheses, weak hypotheses and unknowns (low certainty). The exercise is meant for decision-makers, key stakeholders, subject matter experts, researchers and analysts, cross-functional reps, and local team members.

To know what aspects of culturalisation are important for a business, Chui Chui introduced us to The Culture Activation Toolkit, which consists of 27 cultural aspects that researchers, product managers, growth teams can use to explore and identify cultural elements that are relevant to their specific challenge, market and domain.

By combining Culture + Strategy + People, you turn insights and local knowledge into practical plans for design, marketing and overall customer experience, ensuring your product meets the needs of your target markets.

Accessibility Operations by Devon Persing

Devon discussing with attendees during the workshop

Author of The Accessibility Operations Guidebook, Devon started off by doing a quick introduction to accessibility operations.

Accessibility operations aim to address key challenges in traditional accessibility programs, which often carry the responsibility for ensuring product accessibility, overseeing training, tools, and testing, and responding reactively to regulations — usually with priorities set by leadership or legal teams. However, these programs frequently lack the context, authority, and resources needed to drive meaningful, lasting change.

A significant portion of accessibility work is under-resourced, with only 39% of practitioners feeling adequately supported. Common barriers include limited time (56%), inadequate processes (47%), budget constraints (42%), and lack of leadership and team buy-in, as well as gaps in mentorship and coaching.

Devon defines accessibility operations as the philosophy and practice of connecting people, technology, and processes to create, maintain, and improve accessibility for products and services.

The goals of accessibility operations is to build affinity between people (find common cause and grow trust by working and learning together) and enable systems (reduce friction by helping people help themselves).

Maturity and readiness in accessibility are closely tied to both organisational culture and climate. Culture is shaped by leadership and reflected in the organisation’s official missions, values, policies, and processes. In contrast, climate is driven by the broader workforce and represents the unofficial norms, values, and everyday practices. While culture sets the direction, climate determines how work actually gets done, making both essential to understanding an organisation’s true accessibility maturity and readiness.

After understanding the concept, it was time to evaluate where our organisation is on its accessibility journey by answering prompts and assigning scores for each section:

A. Effort — What is your organisation currently doing around accessibility?

B. Leadership — What does your leadership believe about accessibility?

C. Climate — What do most people in your organisation think and do about accessibility?

D. Knowledge and literacy — What do people know, and how are they enabled to make good decisions?

The score corresponded to a Accessibility Readiness Stage, that ranges from Stage 1 — No awareness, there is no awareness of accessibility, to Stage 7 — Ownership, the community is confident doing accessibility work, using resources, and engaging with each other. Data is evaluated and used to guide new directions in accessibility work. With the Accessibility Readiness Guide we can find a description of each stage and spark ideas for our next steps.

Towards the end of the workshop, Devon presented this interesting concept of communities of practice. Communities of practice are organic, self-selected groups formed around shared interests in areas like advocacy, technology, and policy. They thrive on relationships, mutual accountability, and shared practices — such as stories, tools, and collaborative events — and offer a space for members to seek practical solutions, allies, and support. These communities provide valuable insight into how teams truly operate, serve as platforms for informal learning and experimentation, and help surface patterns and pain points. By fostering trust and grounding solutions in real needs, they become powerful incubators for meaningful, sustainable change.

24 MAY · TALK’S DAY

Revisit the key takeaways if you were there — or, if you couldn’t make it, catch a glimpse of the conversations shaping the future of UX 👉 “UXLx 2025 — Wrap Up — Talks Day”.

👀 Stay tuned!

We’ll start releasing the videos from the Talks Day on our UXLx Videos page, gradually, when we launch the next UXLx event. In the meantime you can check hours of content from the previous editions.

📩 Sign-up to our mailing list to get the latest UXLx news and updates.

🙌 Last but not least, we’d like to thank…

• … our Gold Sponsor Idea Spaces, and our Partners for their support.
• … our incredible speakers who so passionately shared their knowledge.
• … the hundreds of attendees from all around the world who chose UXLx to help expand their knowledge.
• … our photographer José Goulão for brilliantly capturing the essence of our event.
• … the entire UXLx team who always goes all out to bring the best UX content to sunny Lisbon and give everyone an incredible experience.