Snapify

Summary

Snapify is an AI-powered event photography platform that needed to go from a technically capable but disorganized MVP to a market-ready product. I joined as the sole designer, Head of Design, and restructured the entire product: I defined three distinct user roles and their permissions, built a design system from scratch, designed a web dashboard with an AI editing studio, and shipped a mobile app for event guests in time for an investor demo. When I arrived, there was no design documentation, no system architecture, and no established process. When I left, there was a complete product with clear flows for photographers, event hosts, and guests.

Problem Framing

What they thought the problem was: "The dashboard needs better UI."

What the actual problem was: there was no product architecture. The team had built features without defining who uses what, when, and why.

Architecture gaps

• No defined user roles. Three distinct users (photographers, event hosts, guests) were using the same interface with no permission model.
• No settings page, no personal profile page.
• No business model definition: credits vs. monthly subscription vs. free trial?
• No zero states. What happens with 0 albums? 1 album? 200 photos?

Flow gaps

• The photographer workflow was unclear: when does the AI edit? How do presets connect to Lightroom?
• The host approval flow was undefined: a photographer sends photos for approval, then what?
• Guest access was broken: can a guest also be an event host? Can a photographer be a guest at another event?

Design gaps

• No design system.
• Menus were chaotic. The side menu was non-functional and mixed with specification decisions.
• The editing menu was confusing.
• Component sizing was inconsistent.
• The color palette didn't serve the product goals.
• No loading indicators while the AI was processing.

This wasn't a UI problem. It was a product architecture problem wearing a UI mask.

Role and Team

My role: Head of Design, embedded day-to-day as part of the core team. I was responsible for UX research, product architecture, UX/UI design across web and mobile, and visual identity recommendations.

The team:

• CEO / Founder (product vision, investor relations): speed to market, investor demo readiness.
• Lead Developer (backend architecture, AI pipeline): technical feasibility, performance.
• QA (quality, edge cases): stability, error handling.

How I operated: I was the only designer, but I worked there daily like any other member of the team. My job wasn't just to "make it pretty." It was to sit with the CEO and make product decisions together: who are the users? What are the permissions? What happens in edge cases? What's the business model? These are product decisions, not design decisions. But nobody else was asking the questions.

Approach

Step 1: Audit first, design later.

Before touching any screen, I audited the existing interface for systemic problems. The list was long: missing pages, undefined flows, no permission model, no design system, inconsistent patterns.

Step 2: Define the architecture.

I mapped out three distinct user experiences with different permissions:

• Photographers: full dashboard with AI studio, preset management, album creation, payment.
• Event hosts: approval workflow for reviewing and releasing photos.
• Guests: mobile-first experience with QR code entry, face recognition opt-in, and photo sharing.

For each user type I defined: what screens do they see? What actions can they take? What's hidden from them? What happens when roles overlap?

Step 3: Research with real photographers.

I conducted user research with professional photographers to understand their actual workflow. The key insight: they need Lightroom-style editing, not simplified sliders.

Step 4: Facilitate team alignment.

I ran internal workshops to get the CEO, developer, and QA aligned on the same mental model. Many "design" problems were actually disagreements about product direction that had never been resolved.

Step 5: Design system first, screens second.

I built the design system before designing individual screens. Components, colors, typography, spacing, states. Everything consistent from day one.

Challenges

Challenge 1: the "investor event" pressure.

Mid-project, the CEO announced an investor event. I needed a working mobile app for guests to demo face-recognition photo delivery. This forced a parallel track: keep building the web dashboard and ship a mobile MVP at the same time.

The mobile app had its own unresolved product questions: does the guest need face recognition to get photos? What if they don't trust it yet? What happens when they share a photo on WhatsApp? Who releases the photos? I had to make product decisions on the fly, document them, and get buy-in. All while designing.

Challenge 2: AI Studio scope creep.

The AI editing features kept growing: blur detection (but what if the blur is intentional art?), color correction, rotation, duplicate grouping (sometimes 200 near-identical photos), split-screen comparison. Each feature raised its own UX questions.

I scoped it by designing a dedicated AI Studio screen inspired by Lightroom. Professional, powerful, but contained. Instead of scattering AI features across the dashboard, everything lived in one focused environment.

Challenge 3: no foundation.

When I joined, there was nothing. No component library, no spacing rules, no color rationale, no documented decisions. Every screen I designed also had to establish a pattern for future screens.

I was building the foundation and the house at the same time.

Solution

Web Dashboard: Photographers

• Album management with clear states (empty, single, multiple).
• AI Studio: a Lightroom-inspired editing environment with batch presets, blur detection, duplicate grouping, and split-screen comparison.
• Payment integration (Stripe) with a subscription / credits model.
• Settings and profile pages (previously missing).
• Clear indicators when the AI is processing.

Web Dashboard: Event Hosts

• Streamlined approval workflow: upload, review, approve, release.
• Clear status indicators for each photo.
• Defined behavior for rejected photos.

Mobile App: Guests

• QR code scan at the event entrance for instant access.
• Face recognition opt-in, not forced. It can be enabled later in settings.
• Real-time notifications for new photos.
• Favorites, download to device, and share to social / WhatsApp.
• Privacy controls: the user approves photo sharing before others can see it.

Design System

• Built from scratch: components, colors, sizing, zero states, loading states, error states.

Results and Impact

Product architecture

• Defined a complete permission model for 3 user types where none existed.
• Resolved 15+ open product questions (business model, role overlaps, privacy, sharing).
• Created design documentation and a system the team could build on.

Design output

• A complete web dashboard (photographer and host experiences).
• A mobile app (guest experience) shipped in time for the investor event.
• A design system with components, states, and patterns.
• A marketing landing page (designed; not currently live).

Business impact

• The product went from a chaotic MVP to a structured, buildable product.
• The mobile app shipped on the investor-event timeline.
• A unified visual identity across web and mobile.

An honest note on scope: the mobile guest experience was built for an investor demo. The photographer dashboard was used in production. The guest-facing app may not have been fully demoed with real guests at the event.

Next Steps

What I'd do differently

• Push harder for a business model definition in week 1. Payment and pricing decisions affected the UX throughout, and locking them early would have saved rework.
• Bring in structured user research earlier. The Lightroom-style expectation from photographers could have surfaced sooner.

What the product still needs

• A full brand book. The logo and visual identity need professional treatment.
• A marketing strategy. There is no social presence and no investor-deck templates yet.
• Defined customer segments beyond "photographers."

What this project taught me

When someone says "we need better UI," the real problem is almost never UI. It's architecture. It's undefined roles. It's product questions nobody asked. The design work starts before any screen is opened.