AI-Powered Low-Code Drives Digital Transformation

Agents & AI Are Redefining the Future of UX Design Over the past year, AI has become deeply embedded in my day‑to‑day work as a UX practitioner and design leader and the shift is no longer incremental, it’s transformational. Tools like Figma Make, FigJam AI, and Copilot have fundamentally changed how I move from problem to prototype, insight to action, and idea to validation. - Figma Make has redefined fast iteration turning static designs into interactive concepts in minutes and amplifying creative exploration. - FigJam AI has become a quiet productivity powerhouse, accelerating workshops, synthesis, and alignment so I can focus on facilitation and insight. - Copilot supports everything from early research and synthesis to clearer leadership communication, reducing cognitive load across the board. What excites me most is the rise of agentic workflows. I am currently building specialized agents one for interview synthesis and another for unmoderated concept feedback unlocking faster, more structured qualitative insights than traditional methods. AI isn’t replacing designers. It’s amplifying our creativity and impact. But it also raises the bar for us to keep learning, adapting, and evolving alongside the tools. The future of design belongs to those who stay curious and embrace the shift. #agentsinux #AIUX #AgenticAI # AIroleinUX #AIrole #userexperiencedesign #UXdesign #AIinresearch #Userresearch #datasynthesis #Figmamake #Figjam #copilot

The Architecture of AI-Native Experience Designing Decision Environments, Not Just Screens This article draws on ideas from my upcoming book AI Native UX. ————————— For two decades, UX has largely focused on screens — optimizing journeys, reducing friction, refining interaction flows. That work shaped an era of digital products. But that frame is no longer sufficient. When intelligence becomes embedded inside a product — not layered on top — the interface stops being the center of gravity. The real experience is no longer what users click. It is what the system decides. We are moving from interaction-led systems to intelligence-led systems. In AI-native products, users are not navigating fixed architectures. They are engaging with systems that interpret, infer, predict, and act on their behalf. The interface becomes the visible edge of something deeper: a decision environment. In AI Native UX, I describe AI-native products as composed of five foundational layers: • Data layer – behavioral infrastructure. Poor data doesn’t create bad UI. It creates bad outcomes. • Model layer – probabilistic engines that shape trust through their behavior. • Autonomy layer – where the system sits on the spectrum from suggestion to execution. • Escalation & governance layer – how the system handles uncertainty, risk, and oversight. • Interface layer – the expression surface that makes intelligence legible. Understanding these layers is not a technical exercise. It is a strategic one. Experience in AI-native systems becomes the orchestration of intelligence — coordinating models, data, autonomy, and governance into coherent, trustworthy behavior. Screen logic becomes downstream of system logic. This shift has implications for both designers and product leaders. Designers must evolve from interface stylists to behavioral system architects — shaping how systems act, not just how they look. Product leaders must move from feature roadmaps to intelligence roadmaps — deliberately maturing autonomy, data quality, and governance over time. The most powerful products of the next decade will not be distinguished by clean UI alone. They will be distinguished by the coherence of their intelligence architecture. Interface-only UX is not disappearing. But it is no longer sufficient. The real design work now lives beneath the screen. Read the full article here: https://lnkd.in/dSR7ZVRQ #AINativeUX #UserExperience #UXDesign #AIUX #AI #ArfitificalIntelligence

UXtea☕ #06 – Why You Need to Understand AI Product Design #NOW Products are evolving faster than their interfaces. Search, shopping, banking, work tools - everything is getting an AI layer quietly slipped in. And if we don’t understand what AI Product Design really means, we’ll keep shipping “AI features” that feel smart… but don’t actually help anyone. So… what is AI Product Design? In simple words: 👉 #AIProductDesign is about designing how AI behaves - not just how it looks on screen. As designers, we decide: #When the AI should step in #What it should (and shouldn’t) do #How it explains itself #How the user stays in control AI is pushing us from static screens to dynamic, evolving experiences - where we design intelligence, not just interfaces. Most companies want “AI in the product” - not tomorrow… right now. If we don’t design that intelligence intentionally, we end up with features that confuse, overstep, or feel unpredictable. Think about #GooglePhotos. You snap ten similar pictures, and it quietly highlights the “Best Shot.” Not randomly - thoughtfully. It checks smiles, open eyes, clarity, and timing. It shows the suggestion gently, not forcefully. And you’re always in control - accept it, ignore it, choose another. That’s AI Product Design: timing, trust, clarity, and user control working together. #DesignerTip - The 4 Questions to Start Every AI Feature 1. Value — What job is the AI actually doing for the user? Reduce effort? Remove repetition? Save time? 2. Timing — When should the AI act? On request? As a suggestion? Automatically? 3. Control — How does the user stay the boss? Clear options to view, edit, undo, or turn off. 4. Clarity — How do we show what the AI did and why? Simple, human explanations like: “Chosen because this photo is sharper and everyone’s eyes are open.” Not confusing metrics or technical jargon. When these four are done right, the UI becomes effortless. And here’s the truth: The smarter our products become, the more intentional our design must be. That’s where AI Product Design truly matters. #UXtea #AIProductDesign #UXDesign #ProductDesign #AIUX #DesignThinking #DailyUX #MicroLearning #In60Seconds

For 60 years, we’ve been telling computers how to do things. The AI era is flipping the script. Welcome to the first truly new UI paradigm in decades: Intent-Based Interaction. 🚀 Jakob Nielsen recently published a brilliant piece on "Intent UX," and it is a must-read for any leader building digital products today. For decades, we have relied on command-based systems. We navigated complex menus, clicked specific buttons, and executed step-by-step workflows. We had to learn the machine's language. Now? We are shifting entirely to Intent-Based Outcome Specification. Here is the fundamental difference: ❌ The Old Way (Command): You explicitly tell the system how to achieve a goal. ✅ The New Way (Intent): You tell the AI what you want accomplished, and it figures out the rest. We are no longer just "users" navigating an interface. We are directors delegating to an intelligent system. The AI interprets our intent, schedules actions, handles exceptions, and dynamically generates the exact UI we need in that moment. But here is the real C-level takeaway: The companies that win the next decade won't just slap a basic text chatbot onto their legacy platforms. The true winners will build hybrid interfaces - allowing users to seamlessly express their intent while still offering traditional, intuitive UI elements to refine, tweak, and course-correct the AI's output. Are your product and engineering teams still designing for "commands," or are they actively building for "intent"? I’d love to hear your thoughts on this shift below! 👇 #ArtificialIntelligence #UXDesign #Leadership #Innovation #IntentUX #FutureOfTech

Most Design Systems have traditionally been built for software, focusing on screens and ensuring consistent structure and predictability. However, my recent exploration into creating a Design System for an AI platform has revealed that this approach must evolve. It’s not merely a design system anymore; it’s an experience architecture for intelligence. Design Systems have excelled at providing predictable interfaces where buttons behave consistently, layouts respond, and components fit together seamlessly, much like LEGO pieces. Yet, as we integrate curated intelligence into these experiences, we must rethink how Design Systems adapt. In conventional software, the interface is the experience, presenting a set of predefined options within a controlled environment. Even dynamic systems remain bounded, offering users choices. In contrast, AI fundamentally alters this relationship. The interface no longer just presents options; it generates them, transforming the design challenge from creating interfaces between humans and software to crafting interfaces between humans and intelligence. This shift redefines the role of design. It’s no longer just about: - Where things go - How they look - What they do Now, it encompasses: - How intelligence shows up - When it acts versus when it waits - How it earns trust - Where humans maintain control The next generation of systems will be built from behaviors rather than components alone. Instead of asking, “What should this component look like?” we are now inquiring: - When should AI assist versus automate? - How should recommendations be integrated into real workflows? - How can we make machine reasoning understandable rather than magical? - How do we design for trust without hindering efficiency? Currently, there isn’t a mature playbook for this evolution. Rather than simply adding AI to existing components, we are defining a new set of interaction primitives: - Assist - Generate - Recommend - Explain - Simulate - Automate These are not just features; they represent the grammar of intelligent interfaces, outlining how a system communicates its thought process. In this context, the design system. #AIUX #AIDesignSystems #AIProductDesign #DesignSystems #AgenticAI #UXLeadership #AIPlatforms

Are you a UX Designer, or a Prompt Operator? 🤖✨ As generative AI becomes a standard fixture in our workflows, a vital distinction is emerging in our field between operating a prompt and truly designing a user experience. Currently, it is quite common to leverage tools to generate impressive user interfaces—perhaps a sophisticated fintech dashboard—with just a few well-crafted text prompts. While the output is often visually striking, stopping at this layer presents a long-term limitation for our growth as UX designers (in my humble opinion). It’s a bit like slapping an HDR filter on a smartphone pic in Snapseed and believing you're a master photographer (been there, done that 😅). Over the years, the one thing I've understood as a UX designer is that the core mission hasn't changed. Our job isn't asking, "What can AI produce for me?" It's asking, "How can AI shape better user experiences?" Let's take that Fintech dashboard. Instead of just generating static charts to show a user they overspent, what if we used AI to solve the actual user problem? 💡 Anticipatory Design: Using AI to teach financial discipline through "smart friction" before a bad purchase happens. 💡 Contextual UI: Morphing the dashboard's information architecture based on the user's immediate context (e.g., elevating "Rent & Utilities" on the 1st of the month). 💡 A New Metric: Shifting from tracking "Time on Task" to measuring "Prevented Regret." Using AI to generate a UI is just painting the surface. Using AI to build financial literacy, efficiency, and healthier user habits, isn't that beautiful UX? Reminds me of the good old "lipstick on a pig" analogy. How are you moving past the "Prompt Operator" phase in your own design work? I'd love to read your thoughts in the comments. 👇

When AI generates the UI, who's responsible for the experience? That question came up on a real project — a platform where engineers work with complex, variable data. A fixed interface couldn't handle the range of contexts they operated in. So the product direction moved toward letting AI assemble the UI at runtime. Which immediately created a UX problem. If AI is generating the screens, what stops it from hallucinating layouts? Inventing new navigation? Changing the interface every time a user refreshes? All the things that destroy trust in a product. I spent time working through this and wrote a framework for it. The core idea is simple: AI should orchestrate, not create. That means the AI doesn't design. It selects — from a pre-approved set of components, page archetypes, and themes — and assembles them based on user context and intent. The UX team defines what's available. The AI picks and arranges. A few principles that came out of this work: → Predictability over intelligence — same user, same role, same intent should produce the same UI. No surprises. → Declarative output only — AI generates structured schema, never executable UI code. → Page archetypes — AI can only refer to a finite set of pre-defined page structures (dashboard, list, summary, comparison etc). It cannot invent new ones. → Fallback by default — if AI fails to generate within the guardrails, a stable default UI is always available. → Explainability — users should understand why the screen looks the way it does. "Based on your role and recent activity, this screen was generated" — that kind of transparency reduces anxiety and builds trust. This came from a real problem on a real platform. The sketch on my notepad from January is still sitting on my desk. GenUI is coming to enterprise products whether teams are ready for it or not. The design question isn't whether to allow it — it's how to control it without killing the flexibility that made it worth doing. #GenUI #GenerativeUI #UXDesign #AIDesign #EnterpriseUX #DesignStrategy Image Credits: Marcin Czyzowski

When Product, Engineering, and UX align from the start, AI accelerates success. Miss one perspective? AI helps you build the wrong thing faster.

Your design system now has a second user: AI. When you’ve built a design system from scratch twice, in fintech, for products with complex domain models, you start to think you know exactly who it’s for. The answer always seemed obvious: designers and developers. Turns out, there’s a new user in the room 🤖 Over the past few weeks I’ve been experimenting with Figma Console MCP by Southleft (TJ Pitre). I connected a design system to Claude, passed tokens, components and their variants, and decided to see what happens when an agent can read the system directly. I expected convenience. I got something more interesting. An agent that understands your design system better than a new designer on the team. It sees connections between components, understands the logic behind tokens, and can extend the system without breaking it. 🧠 This experiment unexpectedly changed the way I think about design systems. Before, a design system was a tool for people. A library of components and rules that a designer opens, picks the right element from, and applies to the interface. If it was well documented, logically structured, and clear to the team, we called it a good system. ⚙️ MCP changes that logic. A design system becomes not just a tool for people, but a context for AI. The agent reads it programmatically, gets real context, and starts working differently. Not “generate a button”, but “generate a button from our system, with the right state, correct tokens, and in the right context”. The difference turns out to be huge. But there is a catch. For this to work well, the system needs to be explicit not only visually, but structurally. Implicit dependencies, logic that exists “by feel”, component variants without clear rules of use, all of this becomes noise that the agent struggles with. At some point I caught myself asking a new question. Before, I thought about how understandable the design system is to a designer. Now there is a second question: how readable is it for a machine? Are the tokens, component structure, and their variants explicit enough for an agent to work with them without losing meaning? Most design systems are not ready for this yet. And most designers are not thinking about it as their responsibility. ❓ Has anyone tried connecting their system to AI tools? What broke first? #DesignSystems #ux #ai

30 Days. 50+ AI Tools. $$$ Burned. And I finally built a full AI-driven UX workflow — from BRD analysis and research synthesis to IA, scalable design systems, UI, testing, and developer handoff. The biggest shift? I stopped using AI to generate screens and started using it to optimize the entire workflow. AI can easily save 50% of repetitive UX effort. You don’t start from scratch anymore. Ideation is faster. Documentation is clearer. Iteration cycles shrink dramatically. But here’s the truth: AI is acceleration, not innovation. Now that anyone can generate UI, shallow design thinking gets exposed quickly. The real differentiator isn’t how fast you produce screens — it’s how well you articulate design decisions. Why this flow? Why this interaction? Why is this experience better? Without strong foundations in human behavior, usability, accessibility, and core design principles, you can’t properly evaluate AI-generated output. You can only accept it. AI isn’t lowering the bar for designers. It’s raising it. For years we designed guided flows. Now we’re designing systems where users shape the experience themselves — or delegate it to agents. Not fixed journeys, but generative environments. That’s the shift. And honestly, it’s the most exciting one yet. If you’re exploring AI in UX, happy to connect and exchange ideas. Let’s build smarter — not just faster. 🚀 #UXDesign #AI #ProductDesign #DesignSystems