Finde*

Building a marketplace from zero to launch.

*Finde is a vintage fashion marketplace built independently on Django, Python, and AWS. Integrates the eBay enterprise API, S3 image storage, Celery async processing, and a computer vision pipeline for automated garment classification.

PRODUCT

Secondhand shopping is inherently exploratory, but most platforms treat it like keyword search. Finde is a vintage marketplace built around that gap; structured enough to filter in real time, open enough to browse serendipitously.

ROLE

Founder, Product Designer & Builder

PROBLEM

Secondhand marketplaces are fragmented by default: categories are unreliable, inventory quality varies wildly, and trust breaks down before purchase intent can form. Discovery is not just a browsing problem. It is a trust problem, and trust at scale is a growth problem.

SCOPE

End-to-end build: concept, brand, UI/UX, and full-stack engineering.

Finde desktop browse

Wireframe

Onsite

THE BET

Prove that discovery drives buyer engagement. Then use that demand signal to attract sellers.

STRATEGY

I bootstrapped supply through eBay's developer API, launching Finde with real structured inventory from day one. That gave me a live demand signal to validate buyer intent before attracting sellers & building out the supply side.

eBay's developers approved Finde,
granting privileged enterprise level data access
(10K/Day API Rate Call Limit)

BRAND

The brand identity system was built around warmth and familiarity. A serif wordmark, an earthy palette with pops of color, and a custom icon set that made the system feel cohesive.

SYSTEM

Typography, color, iconography, UI and motion were designed together from the start, so the brand scaled consistently across all touchpoints.

finde Rare vintage, all yours.
Primary Secondary Inactive
Linen Vacation Date Night
Designer Wedding Guest
Dropdown Filter
Search... Sort
Size x x x x Size x x x x x x x x x x x x
USER RESEARCH

Early user testing proved that emotional browsing happens before filtering. In response to user research, design focused on discovery first. Curated toggles like 'Date Night' emulate how buying intent forms by season & occasion.

FILTER SYSTEM

Most marketplaces search within a rigid category. Finde searches across all of them. Enter your size & keyword, and cross-category items are returned together in one tailored view, built around outfits rather than isolated items.

PROCESS

Each iteration of the filter system design was pressure tested against competitive references and specific usability criteria. The final design: horizontal dynamic tabs for fast context switching across the full viewport. UI/UX is consistently effortless, across all devices.

ARCHITECTURE

Secondhand browsing is high-volume and low-commitment, so search results update instantly, without page reload, unlike competitors. The backend was built around this foundational decision.

IMPACT

With an initial beta release, organic users reached 28 countries. Session data confirmed high-intent behavior: successful filter and search interactions.

WHAT'S NEXT

The next phase deepens engagement through editorialized lifestyle content, while direct seller outreach informs onboarding tools to grow the supply side. Invite-only curation keeps quality high as the marketplace scales.

Remote testing

Use your keyboard arrows