Designing in the Age of AI – Are We Redundant?
At Versatile.ai, we use data and AI to help the construction industry build smarter and safer.
After eight months here, I thought it was a good time to reflect on how our UX team has been growing and structuring itself — adapting to the specific needs of the company, and to the new generation of AI tools that make our work smarter and safer.
A key component of our mission is to increase velocity and agility without losing clarity, quality, or user focus.
This post isn’t about a perfect method — I don’t think such a thing exists. Methods evolve as tools evolve.
This is about the robust, adaptable framework we’re building to support innovation, collaboration, and growth.
In the past two years, we’ve all been talking about what AI might replace.
But maybe the real question is: what should stay human?
My Take
AI doesn’t replace human thinking — it reveals what kind of thinking matters most. It exposes the gaps in how we work, how we communicate, and how we make sense of complexity.
I use ChatGPT or Gemini to ideate, brainstorm, structure ideas, sharpen tone, explore alternatives, and broaden scope. PMs use Claude Code to prototype live during discussions — it helps them translate thoughts into tangible flows. We experiment with Figma Make to deliver faster, connecting design, documentation, and systems in smarter ways while keeping control in our hands.
But none of these tools make decisions.
They accelerate reflection and execution, helping us see problems more holistically — and achieve better results faster.
Still, the results are ours.
My Rule for AI
If it replaces a task, great. If it replaces a conversation, we’ve gone too far.
That said, it’s an excellent tool to prepare for conversations — to make every meeting more meaningful.
AI helps us move faster, but moving faster only matters when you know where you’re going.
That’s why I keep focusing on methodology and process — the parts of our craft that technology can accelerate, but never define.
A Structure Built for Fast Pace, Clarity, and Collaboration
Our UX/UI team belongs to Product, but our work lives inside the same Jira project as R&D — through a dedicated UX board and a new UX/UI task type we built together.
In previous companies, design often had its own Jira project and workflow. Here, we took a different route — a shared space where Product, Design, and Development collaborate transparently. It’s not just efficient — it’s aligning.
This hybrid setup keeps everyone connected to the same flow of priorities and progress.
We use Slack for daily syncs, Notion (driven by PMs) for product research, and FigJam for UX wireframes and user journeys.
Our Figma Design System, organized with design tokens, ties it all together — and I’m currently improving how our final deliverables in Figma integrate into this structure to keep the same level of quality through every phase.
Every tool and process we refine serves the same purpose:
➡️ Deliver value faster, with fewer frictions, and with more visibility for everyone involved.
Listening Before Changing
None of this happened in isolation.
Before proposing anything, I spent time listening — one-on-one with PMs, developers, QA, ConOps, Sales/CS, and IT. Understanding how things actually work matters more than designing a new system on paper.
Every improvement grows from discussion, not direction, and from the collective intelligence of the people who make things happen every day.
We’re a two-person UX team (soon three — I’m hiring!), and we move forward thanks to a culture of openness shared across Product and R&D.
Our Rhythm
- Kickoff – PM shares context, scope, and hypotheses.
- UX Ideation – short loops, shared exploration.
- Unveiling – team review focused on flow and logic, not visuals.
- UI & Delivery – final designs and components in Figma for Dev and QA.
- Design QA – a quick review before technical QA.
Between these steps: async updates, shared notes, and short syncs — enough structure to stay aligned, not slowed down.
This rhythm helps us keep momentum while maintaining quality — a balance that matters deeply in a startup like ours.
🎯 What I’ve Learned So Far
At Versatile.ai, where change is constant and challenges evolve daily, the values that make the difference — at least to me — are transparency, humility, and adaptability.
I don’t pretend to reinvent the wheel. But I do believe the wheel needs adjustments.
We’ve created a work environment where clarity leads to speed, and where strong processes serve people — and ultimately, our users.
Because at the end of the day, every iteration, every tool, every improvement has the same purpose:
➡️ to better understand, support, and empower the people who use our product.
We create faster with AI. But we don’t create better without curiosity, empathy, and shared purpose.
And maybe the next challenge for us is this 👇
How do we make the last step — the design deliverables, the handoffs, the QA — as fluid and alive as the rest of the process?
Curious how other teams are approaching this.
That’s where the next iteration begins.
💬 Next time, I’ll share more about how we connect this internal process to our users — from interviews to data insights — and how that feedback loop keeps us grounded.
Writing this post was almost as iterative as our design process 😅
Huge thanks to Yigal Carmi (VP Product), Eyal Gersht (VP R&D), Eishay Smith (our CEO), and my dear friend Philippe Adjiman — for taking the time to read, challenge, and refine my thoughts with patience and kindness.
You made it much better than it would’ve been on my own.