How to design AI products that people actually use — and trust. Most AI features fail quietly. They’re generic, verbose and riddled with subtle hallucinations that users can’t easily spot.
Content:
So people ping-pong through prompts, lose confidence in outputs, and eventually stop using AI features altogether. The problem isn’t AI — it's the design and UX around it.
In this 2-days workshop, Vitaly Friedman (Sr. UX advisor with the European Parliament, UX Lead, creative lead at Smashing Magazine) cuts through the noise an AI hype with practical design patterns for AI experiences people actually understand, value and trust — with practical UX guidelines and tons of real-life examples.
All workshop topics:
• State of AI 2026, from limitations to capabilities
• Where companies fail and succeed with their AI initiatives,
• Recent UX research on how and when people use AI features,
• Design patterns for AI products that actually work
• Context engineering and how to design for it
• Agentic AI, and how to design boundaries and guardrails
• How to design for trust and confidence for AI
• How to prioritize impactful AI features
• Launching AI features, where teams fail/succeed
• AI-friendly design systems, and how to set them up
• 100s of real-life examples and UX guidelines
We’ll explore how we can help users get work done better and faster — with better context engineering, designing for building and maintaining trust, applying the right guardrails at the right time for the right tasks and use plenty of emerging design patterns like style lenses, prompt constructors and feedback knobs.
Bonus: We’ll also explore how to adapt design systems to be AI-friendly and how to adjust the workflow to produce better AI-powered results.
Format
This isn't a passive “talk”-workshop. Throughout both days, we’ll switch between short focused sessions and hands-on group work — you’ll apply patterns and techniques to real UX challenges right away, with feedback from the group and from Vitaly. You’ll leave with a toolbox of practical techniques and patterns ready to use next day.
Course discription
The State of AI UX in 2026
• What AI can and can’t do reliably, and what should we as designers and engineers understand about AI accuracy and AI capabilities to create compelling AI experiences,
• New UX research on how people discover and use AI features, and why many AI features have poor adoption and poor retention rates
• How people work with and around AI features, with main slowdowns, blockers, usability and (often overlooked) severe accessibility issues
• AI constraints, limitations and blind spots that most teams overlook entirely
Design Patterns That Actually Work
• Quiet AI vs. Visible AI, prompt strength indicator, daemons, modifiers, task builder, forced ranking, style lenses, precision knobs.
• How to label and signal AI to meet AI compliance regulations and set right expectations with the users,
• How to reduce high interaction cost of prompt engineering, context awareness, capability awareness and AI discoverability
• Hands-on design patterns: style lenses, prompt constructors, precision knobs, task builders, forced ranking, presets, pre-prompts, queuing
• Design patterns for building trust and confidence, AI explainability, designing better reasoning traces, proactive vs. reactive AI, navigation and refinement journeys.
• Context engineering, AI Harness, AI evals, and agentic memory — how to design with them, not around them
AI In Complex Products
• Where AI should live in your product — and how to help users provide better context and make sense of AI output
• Guardrails, approval flows and human-in-the-loop: how to support high-stakes decisions without blocking users in complex workflows
• Overreliance vs. aversion: how to find the balance that minimizes irreversible mistakes
Designing For Trust and Confidence
• Design strategies that help build confidence and clarity with AI — without overwhelming users with disclaimers and reasoning traces.
• Reasoning traces, consensus meters, calibrated trust and limited transparency signals that actually improve retention
• How to maintain trust over time with a set-up of guardrails, feedback loops, refinement and customization
Real-World Case Studies
• How to design AI features and experiences from scratch — with AI Design Canvas, from data collection to gathering feedback
• What successful teams do differently, and where even well-resourced teams often struggle
• How to bring AI into the product, from data collection and data cleansing to designing AI loops and user flows.
• How to measure whether your AI features are actually working for your users, and that the UX around these features is good, not just technically functional
The workshop is for product designers, UX leads, UX researchers and product managers who want to feel more confident and more comfortable about shaping, designing, researching and shipping AI features and AI products.
Target audience
This workshop is for open-minded front-end engineers and product designers who’d love to make sense of AI and create AI experiences that people understand, value and trust. No prior AI expertise needed, but please come ready to think critically, challenge assumptions and do actual work.