I can tell you the rationale behind every single one of the projects I have. I also make games.

Wilfrid Laurier's Navigator portal had a usability problem nobody had actually measured — so I measured it.

I evaluated WLU's Navigator portal through a full research cycle: user surveys, a heuristic review against established usability principles, and moderated usability testing sessions.

Findings were synthesized into empathy maps, personas, and journey maps to give stakeholders a concrete picture of where and why users struggled — then compiled into a final research report with prioritized, actionable design recommendations.

Late-night food access on campus was a real, unaddressed pain point — I designed toward a fix, not just a diagnosis.

Starting from user research into campus food-access pain points, I proposed an app concept that integrates food delivery with social features — addressing not just availability, but the isolation of eating alone late at night.

Built a low-fidelity Figma prototype with usability and accessibility principles applied deliberately from the first wireframe, not bolted on afterward.

A dashboard is only useful if the person reading it isn't a data scientist — so I designed for the second kind of reader.

I built and deployed an interactive COVID-19 data dashboard end to end, but the design decisions are what make it usable rather than just functional: a colorblind-safe palette chosen deliberately over default red/green conventions, unified hover interactions that surface exact values without cluttering the chart, and outlier-handling decisions made specifically so a non-technical viewer sees the real trend, not reporting noise.

The goal throughout: communicate insight to someone who has never seen a rolling average before.

Every menu here optimizes for getting the player back into the game fast — except the one place friction is doing them a favor.

The menu system — main menu, options, pause — is built around a simple constraint: keep what's on screen within what a first-time player can hold in mind at once, so nobody has to stop and parse the interface before they can play. The main menu offers exactly one dominant action, Play, with Options and Exit sized down and positioned to minimize cursor travel between them, and a Credits button deliberately tucked into a corner as the lowest-priority, obligatory-but-present element.

One deliberate exception to "minimize friction everywhere": the back button on both the Options and Credits screens sits in the top-left corner, away from where the cursor naturally rests after using the screen. That's not an oversight — the slight extra travel functions as a soft confirmation, distinguishing an intentional exit from an accidental one.

The pause menu is designed for real interruption, not just game-state interruption: audio sliders stay reachable mid-game, since volume complaints happen mid-session rather than before it, and a controls reminder is present so a confused player never has to alt-tab out. Either exit option, resume or main menu, preserves progress, so stepping away costs nothing.

Playing the entire analysis in Arabic surfaced a bug an English playthrough never would have caught.

I played through in Arabic specifically, not English, to evaluate the localization as part of the interface rather than an afterthought. That choice paid off immediately: subtitle lines were rendering in the wrong order, because Arabic's right-to-left script wasn't being handled correctly by the text engine — a real, specific bug that a default English playthrough would never surface.

The feedback design elsewhere is genuinely thoughtful: the skill tree highlights which part of the mech a selected move physically affects — no functional purpose, pure feedback, and it works. More interesting is the soundtrack, which shifts from peaceful to tense to combat-intense based on proximity to danger — audio functioning as a real interaction feedback channel, not just atmosphere.

The clearest failure was the save system: autosave-only, indicated by a small, faded icon easy to miss, with no manual save option. I lost an entire level's progress to it once, confirming a real cost, not just a theoretical one.

A HUD that respects the player enough to disappear when it's not needed — undercut by smaller failures to extend that same respect everywhere.

The main HUD packs genuinely dense information — minimap treasure, enemy, and waypoint markers, party portraits with live HP/MP for tactical decisions, a Magatsuhi gauge, moon phase — without feeling cluttered, and it disappears entirely while moving so the world isn't fighting the interface for attention. Compared directly against Zenless Zone Zero's persistently busy UI, the restraint is the point.

The pause menu's seven submenus are where the cracks show: the demon stock isn't sortable and reorders unpredictably as party members are swapped, and unusable skills and items stay visible in grey rather than being hidden, adding scanning effort for no benefit. Small, specific, fixable friction, not a rethink of the whole system.

One inconsistency stood out on its own: selecting an Xbox controller icon layout in settings doesn't reliably keep Xbox icons on screen — a small trust-breaking bug in an otherwise disciplined system.