Field Notes

My son is obsessed with F1. He lines up his cars on the floor, and before every race I count him in: “One for the money, two for the show, three for Guido, and cuatro to go!” It’s our line from his Pixar Cars races.

I wanted to give him the real F1 feeling instead: five red lights, a pause, lights out. I looked for an app and didn’t find one I liked, so we started using a YouTube video of the start lights. It worked, but it wasn’t really built for this.

So I made a small app that does the basics. It starts the race with the lights and times it until the first car crosses the line. It works, and it’s what we use now.

It’s a tiny experiment that put a big smile on my son’s face.

Try it: https://race-start-lights.vercel.app/. It works on your phone, and you can add it to your home screen to use it offline.

I speak Spanish, but mostly when I have to. It switches on when I'm talking to someone who doesn't speak English, or doesn't speak it well.

At home, we both speak English, so there's little to no Spanish around the house. And repeating everything in both languages feels like a chore.

I want my son to learn Spanish, but not just any Spanish. I want him to speak like his people.

Spanish changes depending on where you're from. Even everyday words are different: a bus is an autobús in Spain, a camión in Mexico, and a guagua in the Dominican Republic.

So I started wondering: could I build an app just for Dominican Spanish?

I know AI can build apps. But can it build a good one? One that's thoughtfully designed and means something, not another generic AI product?

I want to find out how far AI can get on its own, and how much of a human it takes to make something good.

This album cover started with a question: what should New York feel like for Sundate?

I began by gathering client references and defining the world around the musician — the setting, atmosphere, pose, and clothing. From there, I explored rough sketches and alternative poses before narrowing in on the composition.

The first draft was about finding the right balance between the character and the city. A tighter crop brought more attention to the musician, while the wider view kept the New York setting present. Once the line work felt right, I began testing the color direction: a warmer sunset, brighter and punchier tones, and stronger shadows and highlights to give the illustration more depth.

The final cover came together through that back-and-forth process: reference gathering, questions, sketches, composition studies, and color tests. Each stage helped move the piece closer to something that felt personal to the artist.

I struggle and emerge. I started as a traditional artist, working in charcoal and paint. Last December I took on an album cover in a style I don't usually work in. I'm glad I did. Leaning in beat staying comfortable. I want to keep doing that: say yes to the work that stretches me, not just the work I already know how to do.

This is my first note. I'm excited to be closer to launching a site built around what I actually need — and just the act of building it is pushing me to use what I know and pick up what I don't.

Try the live demo →

A creative exploration of sign-up form design, focusing on reducing friction and creating visual delight. This challenge explores gradient-based design, glassmorphism effects, and real-time validation patterns.

Try the live demo →

The question

What if I designed a sign-up form that feels delightful and reduces friction, rather than feeling like a chore?

Hypothesis

By focusing on progressive disclosure, clear visual feedback, and thoughtful micro-interactions, I can create a sign-up experience that users actually enjoy completing.

Process

Started with exploring different visual directions - minimal vs. bold, single-step vs. multi-step. Chose a gradient-based design with glassmorphism effects for a modern, premium feel. Implemented real-time validation with helpful error messages positioned below each field. Added password strength indicator with visual feedback (5 levels: Weak to Very Strong). Tested various animation timings for the success state, settling on a smooth fade-in with scale effect. Used React Hook Form for form state management and Zod for schema validation.

What worked

Gradient backgrounds created visual interest without being distracting. Password strength meter provided clear, immediate feedback. Real-time validation reduced user frustration by catching errors early. Glassmorphism effect added modern, polished feel. Smooth animations made interactions feel premium and responsive. Color-coded password strength made requirements clear.

What didn't work

Initial color palette was too vibrant for some users - toned it down. Password requirements needed clearer communication upfront - added strength indicator. Success state animation was too quick initially - adjusted timing. Some users found the gradient button distracting - could explore solid alternatives.

Key learnings

• Visual feedback is crucial for form confidence
• Password strength indicators reduce abandonment
• Micro-interactions make forms feel more responsive
• Gradient buttons can be effective if used thoughtfully
• Glassmorphism works well for modern, premium feels
• Real-time validation improves UX when done right

Next steps

Try a completely different visual direction (minimal, bold, dark theme). Experiment with multi-step forms for longer sign-up flows. Test with real users for feedback on color choices and interactions. Explore different validation patterns (inline vs. on blur). Consider adding social login options.