A growth narrative, a working framework for manual, AI-powered, and hybrid design, and the full academic case study for the PNW Floor Design mobile app — built to Professor Pak's creative brief template.
This is not a gallery of screens. It is a working academic document built to Professor Pak's creative brief template, expanded with two meta layers a final-year student should own: a growth narrative comparing who I was in 2022 to who I am now, and a working framework for choosing between manual, AI-powered, and hybrid design workflows. The anchor artifact is the PNW Floor Design mobile app — a real product for my real flooring company in Washington, designed using the design thinking process and the Bootstrap component skills from Advanced UX.
In January 2022 Professor Pak wrote a letter of recommendation describing me as "naturally talented and hardworking," with "an incredibly astute attention to detail" and a "methodical creative process." That was the student version of me: careful, observant, willing to refine a single composition until it locked.
Four years and a return to school later, the careful attention is still there. What is new is the system thinking. I no longer try to make one beautiful artifact at a time. I build the rules the artifacts come from.
"Critical thinking is your strongest skill. Build custom design systems to stand out from generic UI components. Strong case studies are the most important strategy."
Min Pak, M.F.A. · Chair of New Media & Digital Design, Pierce College
The honest question every working designer faces in 2026 is no longer whether to use AI. It is when. The framework I work from is simple: do not let AI make a decision that defines the brand or the user. Let AI carry the volume so the designer can hold the judgment.
When the human owns it.
Why manual: these are the choices that compound across every screen. A bad token poisons the whole product.
When volume is the bottleneck.
Why AI: tasks where doing it by hand costs more attention than the work is worth. AI does the boring so the human can do the brave.
When AI proposes, the designer judges.
Why hybrid: a designer with a clear system can use AI without losing themselves in it. A designer without a system gets eaten by it.
The rule I keep: if a decision can be reversed cheaply, AI can take the first pass. If a decision shapes everything downstream, the human takes the first pass.
Every tool in my workflow earns its place by what kind of decision it serves.
| Task | Tool | Mode | Why |
|---|---|---|---|
| Design system authoring | Adobe Dreamweaver · VS Code · CSS variables | Manual | Tokens define every downstream decision. I write them by hand so I know exactly what each one does. |
| Wireframes & screen flow | Figma | Manual | Logic and information architecture are mine. Figma is the canvas. |
| Brand identity, color, type pairing | Figma · Adobe Illustrator | Manual | If the brand is borrowed, the trust is borrowed. I design my own marks. |
| Hero image exploration | ChatGPT image gen · Adobe Firefly | AI-Powered | Eight directions in an hour is impossible by hand. I pick one and refine. |
| First-draft copy | Claude AI | Hybrid | I outline. Claude drafts. I rewrite in my voice. Voice is mine, volume is shared. |
| Production HTML & CSS | Adobe Dreamweaver · Claude AI · Bootstrap 5 | Hybrid | Dreamweaver gives the editor and SFTP. Claude helps with patterns. The architecture decisions are mine. |
| Asset management, batch operations, compression | Python scripts · Adobe Acrobat | AI-Powered | Renaming 29 files, compressing 153MB PDF to 14MB, recursive uploads. Volume tasks. |
| Photo editing, background removal | Adobe Photoshop · remove.bg | Hybrid | AI does the rough cut. I do the cleanup the AI cannot see. |
| Accessibility & readability checks | Claude AI · browser DevTools | AI-Powered | AI catches invisible text, poor contrast, missing alt attributes. I act on what it finds. |
| Final visual judgment | My eye | Manual | Always. The buck stops with the designer. No tool replaces taste. |
Following Professor Pak's eight-section creative brief template, exactly as taught in DDSGN 151. The PNW Floor Design app is the synthesis of ten years inside the flooring industry and four years studying UX/UI.
PNW Floor Design LLC was created from resilience, creativity, business strategy, and a vision to modernize how flooring companies connect with people. I worked directly with customers inside the flooring industry for over ten years — helping homeowners select materials, solve layout problems, understand product performance, and create spaces that feel aligned with their lifestyle.
Over time, I noticed the same gap on every consultation. Most companies focus only on selling product. Very few design for customer experience, digital systems, visual organization, or project transparency. I saw confusion. I saw homeowners overwhelmed. I saw businesses struggling with their own workflows. I realized I could help bridge that gap, but to do it well I needed UX/UI training. That decision is what brought me back to Pierce College in August 2024.
Problem statement. Homeowners struggle to plan renovation projects because they cannot visualize materials, understand the steps, or manage communication with their flooring vendor.
Goals.
Beauty. Quality. Lasting impressions. Flooring that transforms every step.
Most flooring companies sell product. PNW Floor Design sells the experience around it. The app turns a high-anxiety purchase into a guided, transparent process where the homeowner stays in control from material selection through final install.
Two real personas synthesized from ten years of showroom conversations.
Age: 42 · Location: Tacoma, WA
Role: Working parent, second home renovation
Goal: A flooring contractor who keeps her informed without phone calls. No time for showroom visits between work and pickup.
Frustration: The last contractor disappeared after the contract signed. She had to chase status updates.
Needs from PNW app: Project timeline, in-app chat, push notifications for delivery and install dates.
Age: 31 · Location: Seattle, WA
Role: First-time homeowner, never picked flooring before
Goal: Understand what the categories mean (LVP vs. engineered hardwood vs. tile) without sounding ignorant to the designer.
Frustration: Houzz is inspiring but overwhelming. Floor & Decor's site does not explain enough. Home Depot is product-first, not design-first.
Needs from PNW app: Material Library with plain-language descriptions, filters by use-case (bathroom, kitchen, basement), saved favorites.
| Competitor | Strength | Weakness | What PNW does instead |
|---|---|---|---|
| Houzz | Strong inspiration library, large content catalog | Overwhelming. No project tracking. Generic content, not tailored to the homeowner's actual project. | Curated Material Library tied to the homeowner's specific project. Filters by use-case, not by style alone. |
| Floor & Decor | Wide product range, in-store sampling | No AR visualization. No way to preview a sample in your actual room before buying. | Future AR visualization layer. Saved material metadata that survives the showroom visit. |
| Home Depot | Convenience, scale, financing | Product-first, not design-first. No designer relationship after purchase. | In-app chat with the assigned designer. Project visibility from quote through final install. |
Ten years of real showroom conversations turned into a research foundation. Documented the most common homeowner pain points: not remembering samples, losing communication after contract signed, confusion about material categories.
Mode: Manual. Real notes, real customers.
Synthesized findings into two anonymous personas (The Working Parent, The First-Time Homeowner) and a customer journey: discover → consult → select → quote → install → live with. Identified post-contract communication as the steepest emotional drop.
Mode: Hybrid. AI helped cluster interview themes. Persona writing was mine.
Hand sketches and Figma low-fi wireframes for the ten core screens. Focus on flow, not pixels. Confirmed the navigation pattern (bottom-tab) and the priority of Material Library, Project Timeline, Chat.
Mode: Manual. The structure had to be mine.
Locked the visual design system (lavender, paper, dark-green primary, soft pink secondary, SF Pro + Source Serif Pro). Built component library first. Screens compounded instead of fighting each other.
Mode: Hybrid. AI helped with first-pass component variants. I refined every screen.
A 60-page case study PDF documenting every screen, every flow, the persona reasoning, and the visual system. Originally 153 MB. Recompressed to 14 MB without losing readability.
Mode: Hybrid. Manual writing, AI-powered compression.
Final pass on every screen against the design system rules: limited palette, two-font pairing, single type scale, two radius values, subtle hover only. Stripped anything that drifted.
Mode: Manual. The system is the editor. I am the system.
The visual design system for the PNW Floor Design app uses a different palette than this portfolio site (a brand should not borrow from another brand). The app's brand is lavender luxury — soft, premium, calm. Below are the locked tokens.
Soft luxury. Minimal, clean layouts. Human-centered digital experiences. The visual language signals "premium service" without being cold or corporate.
| Token | Family | Size | Use |
|---|---|---|---|
| H1 | SF Pro | 36px | Screen titles |
| H2 | SF Pro | 24px | Section headers |
| H3 | SF Pro | 20px | Subsections, card titles |
| H4 | SF Pro | 18px | Tertiary labels |
| H5 | SF Pro | 16px | UI labels |
| H6 | SF Pro | 14px | Tags, eyebrows |
| Body | Source Serif Pro | 12px | All long-form copy |
Warm. Plain language. Confident without bragging. A friend who happens to know flooring. Never sales-y, never cold.
Custom PNW house mark in deep green. Custom iconography for the three anchor features (tile sample, wood plank stack, heated-floor element). Icons hand-drawn in Inkscape, exported as SVG for crisp rendering at every screen density.
The full information architecture for the PNW Floor Design app. Every screen earns its place by serving a real homeowner task.
Problem. Homeowners cannot remember which sample they liked. Photos in a phone gallery do not preserve product code, square footage, or installation type.
Feature. Saved materials with full metadata. Filter by room type, water resistance, foot traffic, budget.
Why it matters. Removes the "I think it was the gray one" conversation. The sample stays with the customer.
Problem. After the contract signs, communication drops off. Homeowners do not know when materials arrive, when install starts, or what is next.
Feature. Live timeline of every project phase. Push notifications on phase transitions.
Why it matters. Project visibility is trust. Trust is repeat business.
Problem. Smart-home control of radiant heat is fragmented across vendor apps. No one app knows the room layout the floor was installed under.
Feature. IoT integration mapped to the original install plan. Schedule by zone, by day, by occupancy.
Why it matters. The flooring product becomes the platform, not just the floor.
View Live Figma Prototype → Open Case Study PDF Download PDF
Live Figma file (view-only). Last updated as part of DDSGN 215 Project 2 submission.
The PNW Floor Design app uses lavender luxury tokens. This portfolio site uses warm sepia tokens. Different brands, same discipline. The five constraints below apply to every project I touch.
Five colors means I cannot hide behind variety. Two fonts means the type pairing carries the emotion, not a third font. Three sizes means the hierarchy is decided once. Two radius values means every corner agrees. A 3-pixel maximum hover lift means the interaction is felt, not announced. Restraint is the proof that the designer made a choice.
When I have a system, I can think about the user instead of the pixels. The system holds the pixels for me. When I do not have a system, every decision exhausts a budget I cannot afford to spend.
AI can make my voice louder, faster, and reach further. It cannot replace the voice. The moment I let AI hold the brand voice, the brand stops being mine. The discipline is to ask AI for the volume work and never the identity work.
The 500-site audit on vibe-coded portfolios named the same patterns I had been making: purple gradients, em-dashes, glowing hovers, fake testimonials, placeholder names. None of those patterns are bad in isolation. All of them together signal a missing system. I removed every one of them from this site on purpose.
Ten years of helping homeowners pick tile, calm their anxiety, manage delivery problems, and protect their budget was a product design education in disguise. UX/UI school gave me the vocabulary. The job gave me the empathy. The combination is the work.
A different student than the one Pak met in 2022 is writing this. Same attention to detail. Same methodical process. Now built on top of a design system that survives my own mistakes, an academic creative-brief discipline, and a tool framework that uses AI without losing the voice.