Committing to the full design before opening the engine meant implementation was translation, not invention. Every layout decision (hierarchy, spacing, component structure) was made once, not renegotiated every time an element was placed in the engine.
Before opening Figma, the core loops were mapped out in pseudocode and flowcharts: click-to-treat, upgrade purchase, auto-clicker passive income. The GDD and TDD were produced this week too. Having both written first meant the wireframes were built around known behaviour, not guesswork, and each following week had a clear reference rather than ongoing decisions about what to build.
Pseudocode
Game Design Doc
Then into Figma: low-fidelity greybox for layout logic and hierarchy only, then an improved lo-fi pass, then final wireframes with the complete colour palette, typography, and component structure locked in. Visual direction came from Animal Party, Otter Ocean, Pupperazzi, and Coral Island, cozy but not childish, soft enough to feel approachable, clean enough to feel considered. Peach background, pastel pink shop panel, lavender for locked and coming-soon states. A single rounded sans-serif throughout, with weight doing the hierarchy work rather than mixing typefaces. Gameplay references from existing idle clickers informed the pacing logic and upgrade feel before any UE5 work began.
Colour Palette
Style Research
Style Research
Gameplay Research & References
Separating "make it work" from "make it look right" meant the logic was stable before the visuals arrived. Debugging Blueprint and reworking layouts at the same time would have been slower and produced worse results on both dimensions.
Grey placeholder buttons: the dog was initially just a large clickable button. Swapping in the sprite early was a personal call: having a character there made it much easier to reason about what feedback should happen on each pet. A treat counter that incremented, 2–3 upgrade levels, floating "+1" feedback. Ugly, functional, and essential. Getting a working build early also created buffer time that turned out to be needed when sick leave started eating into available weeks.
First Implementation
Dog illustration: Hand drawn cartoon chihuahua by @freepik (freepik.com)
Because the Figma work in Week 1 was thorough, visual implementation in Week 3 was mechanical. The wireframes were the spec; the engine work was execution. That's a much better place to be than making design decisions inside the engine.
The start screen, custom buttons, correct padding and spacing, typography matching the design system, the treat counter in its final position above the dog. The shop doesn't have to be always visible; a dedicated shop button toggles it, so the dog and the core game view stay in focus. Every screen maps directly to a Figma frame from Week 1.
A distributable .exe is a different kind of portfolio artifact than a video of a playable build. Packaging wasn't required by the brief: it was a self-imposed challenge, and worth doing for exactly that reason.
Final additions: Niagara heart particle effects triggered on each pet (first real Niagara use, requiring exploration of the VFX-to-UI plugin), shop popup slide-in animation, disabled button states for unaffordable upgrades, and a LinkedIn icon linked directly to the profile. Each click triggers four simultaneous feedback responses: a floating "+X" counter, a scale pulse on the dog, a cat purr SFX, and Niagara particle hearts. Every layer intentional.
Packaging took approximately seven hours of troubleshooting, three or four Unreal reinstalls, and mentor help for the Niagara plugin installation. GitHub made the whole process survivable: trying aggressive fixes without fear of losing working code is the difference between problem-solving and gambling.
Niagara Plugin
Particles WIP
