CASE STUDY · PRODUCT BUILD

PROGRIZE

A career platform across three connected surfaces: a mobile app for finding jobs, a desktop view for the slower research work, and an AI prep layer that lives inside the product. Concept to live in five weeks.

← Back to all work2024// BUILD SERVICE →

UX DESIGNMOBILE APPWEB PLATFORMAI INTEGRATIONDESIGN SYSTEM

THE PRODUCT

Three surfaces. One product.

A calm mobile app, a desktop research view, and an in-product AI prep layer, all reading from one data model.

app.progrize.com/jobs

// JOB DISCOVERY

Overview

A multi-surface career product for an early-stage founder: a mobile app, a desktop web platform, and an in-product AI prep layer, all reading from one data model. The headline: concept to a launchable product in five weeks, with full IP handed over.

The challenge

Finding the listing is the easy part.

George's pitch was blunt. Every job platform he'd used was either a feed engineered to keep you scrolling or a giant searchable database that dumped the work back on you. Neither got anyone closer to an actual offer. He wanted the part everyone skips: what happens after you find the listing.

That pointed at three surfaces sharing one set of data, each suited to a different headspace. A calm app for browsing on the move. A desktop view for the sit-down research. And prep folded into the product, so getting ready for an interview doesn't mean closing the tab and disappearing into YouTube.

The wishlist ran longer: a community tab for peer advice, referrals for warm intros, an AI CV optimiser that tailors an application to a specific job. A lot of product for a five-week window, which forced the harder conversation about what actually ships first.

What was hard: three surfaces in five weeks meant something had to give. We chose to ship the app and prep layer fully polished and keep community deliberately thin for v1. One feature that works beats three that don't. If we ran it back, we'd lock the AI feedback scope down on day one. That's where the late nights came from: getting the tone right, not the tech.

What we did

Designed a calm, low-pressure mobile app for browsing on the move, cutting the countdown badges and notification nagging most job apps lean on, so people come back.
Built a desktop web platform with side-by-side context (company reviews, an AI match score, long-form detail) as its own layout rather than a reflowed phone screen.
Integrated an in-product AI prep layer (question banks and feedback on practice answers tied to the role you're applying for), so getting ready never means leaving the product.

// PROGRIZE WEB PLATFORM

// PROGRIZE LEARN LAYER

The outcome

Three connected surfaces shipped: mobile app, desktop platform, and an in-product AI prep layer sharing one data model and design system.
AI prep built into the product, surfaced around the role you're applying for rather than bolted on as a separate course.
From brief to live in five weeks, launched on schedule with full IP and Figma source handed over. No lock-in.

Tech & stack

React Native · Next.js · Supabase · AI feedback layer

THE BETS

What we set out to solve.

Four convictions about job-hunting that pointed us away from the engagement-farm playbook.

OVERWHELM, NOT SCARCITY

The problem most job seekers have isn't too few listings. It's too many, with no way to tell which are worth their time. We designed for filtering and confidence, not for volume.

PREP IS THE GAP

Plenty of products help you find a job. Far fewer help you get ready for the interview once you've applied. We treated prep as a core surface, not a content add-on.

CALM BEATS LOUD

Job hunting is already stressful. Our bet was that a quiet, low-pressure interface would keep people coming back longer than a feed engineered for engagement.

ONE PRODUCT, THREE MODES

Discovery, research, and prep are different mindsets, and pretending they're one screen serves none of them well. Each got its own surface, tuned to that moment, all reading from the same data so nothing drifts out of sync.

THE BUILD

How we built it.

Four phases across five weeks. Scope arguments in week one, three finished surfaces by week five.

Week 1

DISCOVERY & SPRINT 0

Week one was mostly arguing about scope. George wanted everything: community, referrals, an AI CV optimiser. Five weeks doesn't hold everything, so we sorted the wishlist into what makes the product feel finished on day one and what can wait. The shape that survived was three surfaces: a phone app for finding jobs, a desktop view for slower research, and a prep layer that lives inside the product. One data model underneath, three jobs on top.

✓ User flow mapping✓ Information architecture✓ Brand direction✓ Sprint backlog

Weeks 2–3

PROGRIZE APP: MOBILE

The app is the calm one. Browsing sits under your thumb, you swipe to shortlist, and signup is short rather than a form marathon. We cut the things most job apps lean on: countdown badges, red dots, notification nagging. Job hunting is stressful enough without a feed trying to spike your cortisol. A tool that stays quiet and gets out of the way earns the next open, which is the only metric that matters here.

✓ Mobile screen designs✓ Interaction prototypes✓ Design system tokens✓ Dev-ready specs

Week 3

PROGRIZE WEB: DESKTOP

Desktop is where people sit down and actually compare. Same data as the app, but the extra width buys side-by-side context: company reviews next to the role, an AI match score, the long-form detail nobody reads on a phone on the bus. Rather than reflow the mobile screens, we treated desktop as its own layout with its own priorities. Same product, different posture.

✓ Desktop layout system✓ Company review module✓ AI match scoring UI✓ Responsive breakpoints

Weeks 4–5

PROGRIZE LEARN: AI LAYER

Prep usually means leaving the app to go study somewhere else, and most people never come back. So we kept it inside: question banks, AI feedback on your practice answers, and progress tied to the role you're applying for, not a generic course. The genuinely hard part wasn't the model, it was the tone. Early feedback read like a stern examiner. We rewrote the prompts until it read like a coach pointing at the one thing to fix next.

✓ AI feedback integration✓ Question bank UX✓ Progress tracking✓ Notification system

THE OUTCOME

What we delivered.

Three connected surfaces, one design system, an in-product AI layer. Brief to live in five weeks.

CONNECTED SURFACES SHIPPED

Mobile app, desktop platform, and an in-product AI prep layer, sharing one data model and one design system.

PREP BUILT INTO THE PRODUCT

Question banks and AI feedback on practice answers, surfaced around the role you're applying for, not bolted on as a separate course.

DESIGN SYSTEM, REUSED EVERYWHERE

Shared tokens and components across all three surfaces, so the product stays consistent as the team keeps building.

5 WKS

FROM BRIEF TO LIVE

Concept to a working, launchable product in five weeks. Launched on schedule.

HANDED OVER

+ Mobile + web + AI surfaces+ AI job discovery & CV tools+ Full IP + source handed over

“They didn't just make the brand — they built trust. Vazgro delivered faster than any agency I'd worked with, and the AI layer they shipped genuinely set us apart.”

GEORGE A.

FOUNDER, PROGRIZE

HAVE A SIMILAR PROJECT?

Let's build yours.

Book a free call Explore BUILD →← View all projects