Mariona Guiu

Figma is where most of my design work lives, from early wireframes to handoff. Component libraries keep things consistent across products, and having everything in one place makes collaboration easier. FigJam I use for workshops and collaborative sessions, mapping flows and gathering feedback in a more open format.

For more realistic prototyping I've been using Figma Make, which has the advantage of staying within the same ecosystem and connecting directly to existing libraries. When I need something that feels more like a live product outside the design file, that's when I reach for Lovable instead.

I use ChatGPT mostly for quick brainstorming, generating copy variations, or reframing a problem when I feel stuck. It's also been useful for refining prompts when I'm working with other AI tools, since how you phrase things makes a big difference in what you get back.

For heavier writing I tend to reach for Claude. UX copy, documentation, research summaries, it handles tone well and rarely needs much editing. It's also become part of how I process user research, helping me turn interview notes into something structured and actionable without spending hours formatting a report.

Lovable has been useful for those moments when a Figma file just won't cut it and I need something that feels real fast. It helps me get to a working prototype without waiting on a development cycle, which is handy for testing feasibility early. I've also started using it for stakeholder presentations when static screens aren't enough to get everyone aligned on where we're heading.

At Vattenfall, Maze is used for unmoderated user testing across teams, and it's something I'm keen to explore further in my own design process. From what I've seen, it collects useful data on task success, navigation paths, and friction points that can help validate design decisions before development begins. Heatmaps and session recordings also seem like a practical way to catch usability issues that internal reviews tend to miss.

Lottielab is my go-to for micro-interactions, loading states, and transitions. It exports in formats developers can plug in directly, which cuts down on a lot of back-and-forth about motion details and keeps the final result close to what I intended.

For more complex animations I've been exploring Rive. It handles animations that react to what users do, which Lottielab isn't really built for. The learning curve is steeper, but for projects where animation is doing real UX work rather than just adding polish, it's worth it.

Lokalise is something I've worked with at Vattenfall to manage translations and keep localisation organised across the product. It syncs with design files which helps catch text overflow and layout issues before they become a problem in other languages. Working on a product that needs to support multiple languages means translation can't be an afterthought, and having a dedicated tool for it makes collaboration with translators a lot smoother.

I've used Google Analytitcs across different products, including during my time at Post-it app and Mattful, to understand how users actually behave once something is live. Seeing where drop-offs happen, which features get used, and how different users move through the product has shaped a lot of my design priorities over the years. I also find it useful when advocating for design changes, having data on conversion rates and engagement makes it easier to build a case with stakeholders rather than relying on instinct alone.