Hello, my name is Noora

Why this project?

Many gardening apps provide generic advice based on country or climate zone. In Nordic countries, where growing seasons are short and weather changes rapidly, hobby gardeners often struggle to know when conditions are actually suitable for planting.

I wanted to explore how hyper-local weather data, AI guidance, and a simple task-based interface could reduce uncertainty and help beginners grow food successfully.

The problem

Home gardeners frequently ask questions like:

• Is it actually warm enough to plant?
• Should I water today if it's going to rain?
• What should I be doing this week?

Existing apps often provide generic calendars, large plant databases, and isolated weather forecasts, but don't connect these into clear, actionable guidance.

Design goals

• Reduce decision fatigue by surfacing only the next important task.
• Personalize recommendations using location and weather.
• Help beginners succeed without requiring gardening expertise.
• Keep the interface calm and uncluttered.

Research & insights

I gathered signal from Nordic gardening forums and communities, existing gardening and weather apps, and my own experience growing food in a short, unpredictable season.

Product strategy

Rather than starting from a feature list, I anchored every decision to a single core user need.

Core user need: "Tell me what I should do today."

Design process

Information architecture

Navigation is organized around three areas: Today, Calendar, and Garden, with Today set as the default landing screen. Everything else is one tap away.

• Today — the day's most important tasks, generated from weather and calendar data.
• Calendar — the hyper-local planting timeline for each crop.
• Garden — the gardener's own plants and their current care status.

Design system

The visual language was designed to feel calm and approachable while supporting long-term daily use. Green tones communicate growth without overwhelming the interface, while generous spacing keeps task-focused screens easy to scan.

• Typography — DM Sans for body text and UI, paired with DM Serif Display for headings.
• Spacing — a relaxed, consistent scale so single-task screens never feel cramped.
• Components — a small set of reusable cards, chips, and inline confirmations rather than modal-heavy flows.
• Tokens — color, spacing, and radius values shared between design and the frontend implementation.
• Accessibility — AA-contrast color pairs, visible focus states, and no color-only status indicators.

Key UX decisions

Defining the "Today" experience

Users open the app because they want to know what they should do right now, not browse information. The home screen prioritizes today's actionable tasks instead of showing all plant information equally. This reduces cognitive load and supports quick daily check-ins.

Onboarding data collection

The app needs enough information to personalize recommendations without making onboarding feel tedious, so I only collected the information required to generate meaningful recommendations:

• Location
• Growing space
• Experience level
• Selected crops

Every question had to directly improve the user's experience. Location, growing conditions and experience level all influence the app's recommendations and guidance.

Automation vs. manual control

I distinguished between information the system could know with confidence and actions only the user could verify. Weather conditions and care recommendations were automated from external data, while plant care actions such as watering remained user-confirmed. This prevented the system from creating inaccurate plant histories while keeping interaction lightweight through one-tap logging with an undo option.

Returning after a long absence

Gardening is seasonal, and users may not open the app for weeks. Rather than overwhelming users with missed notifications, the app recalculates the current state and surfaces the most relevant actions.

Final UI

Reflection

If I continued the project, I would:

• Validate assumptions through user testing with real Nordic gardeners.
• Measure task completion and retention to see whether the daily-task model actually works.
• Refine AI recommendations based on user feedback over a full growing season.
• Expand the growing calendar beyond Nordic climates.

Why this project?

Coastal Paws is a mockup of a real boarding kennel and cattery on the NSW North Coast. I came across their site while thinking about local service businesses whose websites don't reflect the quality of the service behind them, and it stood out as a great candidate for a self-directed redesign: a small, content-heavy site with a genuinely lovely business behind it, held back by a layout that made it hard to find anything.

I wasn't commissioned to do this, I picked it because it gave me a real, messy information architecture problem to solve rather than a clean slate.

Auditing the existing site

Before redesigning anything, I went through the existing site page by page and logged what was actually getting in the visitor's way.

Information architecture

I restructured the site so each page has a single clear job, and the homepage does the work of answering a first-time visitor's most common questions instead of making them dig for it.

• Home — who Coastal Paws is, the primary booking CTA, testimonials, a facilities preview, starting prices, opening hours, and location. Everything needed to decide and get in touch, on one page.
• Kennels & Cattery — dedicated pages for each service, structured consistently: the space itself, daily care, and boarding requirements.
• Prices — detailed pricing information kept separate so homepage doesn't get cluttered.
• In Memory of — a tribute page to pets who've stayed with them over the years, kept as its own space since it serves an emotional purpose rather than a transactional one.

Key UX decisions

Making the CTA impossible to miss

The single most important action on the site, booking, needed to stop competing with everything else on the page. I gave it its own high-contrast button directly in the hero, set against a full-width photo instead of surrounded by text. I also added a final call-to-action just before the footer included the contact information in the footer.

Bringing decision-critical information forward

Hours, location, starting prices, and a facilities overview were previously buried in the cluttered homepage and subpages. On the redesigned homepage, they're all visible without clicking through anywhere. Reception hours and holiday closures sit in a dedicated, high-contrast section, and address, phone, email, and a map are grouped together right below it.

Establishing a repeatable visual hierarchy

Rather than styling each section individually, I built a small set of alternating section patterns. Soft mint and cream backgrounds, a consistent heading scale, and generous whitespace so visitors can tell at a glance where one topic ends and the next begins, without relying on dense borders or heavy dividers.

Consistent structure across Kennels and Cattery

Both service pages follow the same pattern; the space itself, daily care, then boarding requirements so a visitor who's already read one page knows exactly where to find the equivalent information on the other. The starting prices are clearly displayed on each page, and a button to view the full pricing table is included.

Final UI

Reflection

If I continued the project, I would:

• Run the new information architecture past real visitors through a tree test to confirm it matches how they'd actually look for things.
• Validate the CTA placement with real booking data rather than heuristics alone.
• Reach out to the business to see whether any of this could actually inform a future update to their live site.

Introduction

Dungeon Game is a small three-level mobile stealth game I made using Unity and C#. Choose your weapon and make your way through the levels fighting skeletons and collecting coins. With the coins you can buy better weapons.

My goal was to practice making a small game with a couple of levels, and having data persistence through the levels and game sessions. Making the game I learned to utilize different game design patterns and Unity features such as Events and Scriptable Objects.

Play the game on itch.io

Gameplay

Future features

• More levels
• A wider variety of enemies (currently only one type)
• Other collectable objects, such as keys (currently only coins)
• Improvements to animations, VFX, and sound