< All work

GoldChain

GoldChain is a B2B marketplace for traceable, ethically sourced gold. I designed it from field research through a closed pilot, and rebuilt it when the first version didn’t take.

goldchain.pro
Role
Lead Product Designer
Year
2024
Category
Commodity marketplace
Status
Closed pilot

01 / big idea

Make the sale the product, and tracking the receipt.

Everyone downstream wanted proof of where gold came from. The miners who had to create that proof had no reason to. GoldChain works because every participant comes for the trade, and the record of where the gold came from is produced along the way.

BeforeSmuggling rings, fraud, and gold of unknown origin mixed in with clean gold, with nothing a buyer could check before paying.
AfterA marketplace where verified producers sell to verified buyers, and every sale proves where the gold came from.

02 / my role

  • Ran field research and usability sessions with miners, traders, and buyers
  • Designed the information architecture, flows, and visual system
  • Built the component library
  • Built the platform with Claude Code and Codex
  • Organized the closed pilot and ran follow-up interviews
  • 4Roles in one system, producer, exporter, importer, and refiner
  • 1Scan from a QR code to the full chain of custody
  • 2Countries in the closed pilot, the DRC and the UAE

03 / Problem

The gold trade runs on trust gaps

Small-scale miners have no reliable path to legitimate buyers, and buyers can’t prove where gold came from.

Artisanal and small-scale mining produces about a fifth of the world’s gold and supports millions of miners, mostly in sub-Saharan Africa. The chain from mine to market is informal. Refineries and jewelers can’t verify origin before they pay, miners can’t reach fair prices, and smuggled gold mixes in with clean gold.

  • Market failure

    No trust between producers and buyers

    Verification happens informally or not at all. Buyers take the risk or walk away.

  • Regulatory gap

    Rules on paper, not in practice

    Responsible-sourcing guidelines exist, but nothing makes them achievable for a small mining cooperative.

04 / First version

A tracking tool for regulators

The first GoldChain made the ledger the product. Every handoff, from mine to refinery, was entered as its own block.

  1. 01

    Web form

    Collected transaction data for accounting and investors, replacing folders of PDFs.

  2. 02

    Tracking app

    Recorded every handoff from mine to exporter to importer to buyer, built for regulators.

  3. 03

    Field test

    Put it in front of the supply chain to see who would use it.

Gold moves from the mine through a trading center, exporter, and importer to a refinery or jeweler, and each handoff is written as a block on a consortium blockchain
How gold moved, and how each handoff became a block on the ledger

One form for every handoff

Each participant added a block when the gold changed hands, the source, the auditor, the exporter, and so on down the chain. Every block carried the lot, weight, purity, and a signed declaration, and linked to the block before it. It was a good idea. Regulators and buyers got exactly the record they had asked for. What it was missing was a reason for the miners at the start of the chain to fill it in.

Tracking app wireframes, one form per block

05 / Key insight

Compliance is not a reason to log in

The field test split the supply chain in three, and the people who mattered most said no.

  • End buyers

    Engaged

    Loved seeing origin and handling, but they only consume data. They can’t create it.

  • Refineries and traders

    Selectively interested

    Some saw value, mainly those selling into regulated EU and US markets.

  • Miners

    Reluctant

    Extra work with no economic return. Without them, the chain was incomplete.

We cut the miners’ part to two screens and made their accounts free, and they still had no reason to spend five minutes entering data. A tracking tool only works if the people creating the data want to use it. So I turned GoldChain into a marketplace. Producers list gold to reach verified buyers, buyers are protected by escrow, and every sale writes the tracking record that regulators and end buyers wanted all along.

Miners list to sell, not to comply. The compliance follows the commerce.

06 / Decisions

Complexity absorbed, not exposed

Users from rural cooperatives to Dubai trading desks share one product. None of them see a compliance framework, only their next step.

A marketplace people already understand

Noticed
Miners and buyers knew how to shop online, but not how to navigate commodity compliance.
Decided
Borrow the browse-and-buy pattern of a consumer marketplace, and layer in the signals that matter for gold, trust tier, certification, and quantity.
Result
Producers list to sell, and buyers see verification before they see price.
  • Listings with trust and certification up front
    Listings with trust and certification up front
  • Buyer protection on every listing
    Buyer protection on every listing

Verification, one step at a time

Noticed
Showing the full list of documents up front made verification feel impossible for a small cooperative.
Decided
Reveal only what the next step needs, then give one clear review screen before anything is submitted.
Result
Producers finish listings with every document in place, without facing the whole requirement at once.
  • Review and confirm a new listing
    Review and confirm a new listing
  • Seller dashboard and incoming offers
    Seller dashboard and incoming offers

Protect both sides of the deal

Noticed
Sellers and buyers in new trade relationships had little reason to trust each other.
Decided
Negotiate offers in the product with the listing in view, hold payment in escrow, and design the dispute flow before the happy path.
Result
Neither side has to take the other’s word for it, and a problem has a clear path to resolution.
  • Counter an offer with the listing in view
    Counter an offer with the listing in view
  • A structured dispute if delivery doesn’t match
    A structured dispute if delivery doesn’t match

Trust you can scan

Noticed
The tracking record had value beyond compliance. People buying jewelry or bullion want to know it was sourced responsibly.
Decided
Give every settled lot a QR code linked to a public provenance view, written for people, not auditors.
Result
Anyone, down to the end buyer, can see where their gold came from without an account.
  • Every settled lot gets a QR code
    Every settled lot gets a QR code
  • The public view any buyer can scan
    The public view any buyer can scan

One data model, a view for every role

Noticed
Producers, traders, buyers, and admins all touch the same transaction, with different permissions, and a mistake in a cross-border trade is expensive to undo.
Decided
Share one data model and give each role its own view of it. Make every action one tap, but nothing final without a confirmation, and give admins one place to see every verification, order, and flag.
Result
Every role sees what it needs and nothing it shouldn’t, and every admin action is logged.
  • Nothing is final without confirmation
    Nothing is final without confirmation
  • Admin oversight with an audit trail
    Admin oversight with an audit trail

07 / How it works

From listing to settlement

International trade paperwork, simplified into steps a rural mining cooperative can complete.

  1. 01

    List a lot

    Mine documents, weight, purity assay, and origin, with a reference ID for the lot.

  2. 02

    Verify

    Documents checked and the listing shown to approved buyers.

  3. 03

    Make an offer

    Buyers review the record, negotiate, and pay into escrow.

  4. 04

    Confirm delivery

    The buyer checks weight and quality against the assay, with a window for disputes.

  5. 05

    Settle

    Escrow pays out, the record is sealed, and a certificate is issued.

08 / Tradeoffs

What I built, and what I left out

BuiltEscrow on every sale. More friction, but it protects both sides and brings in institutional buyers.

Left outDirect payment between buyer and seller.

BuiltSupply first. Producers and exporters onboarded before buyer features opened.

Left outAn open marketplace during the pilot. Every participant is invited and verified by hand.

BuiltA B2B marketplace for licensed traders and professional buyers.

Left outSelling to consumers, which would have weakened verification and exposed trade pricing.

09 / Status

In a closed pilot

GoldChain is in a closed pilot in the DRC and UAE with one participant in each role, a producer, an exporter, an importer, and a refiner, under government oversight. The pilot tests the trust model before scaling, so I’m not publishing volume numbers until it ends.

10 / takeaway

What I carry forward

The first version was polished and technically sound, and the people at the start of the chain simply ignored it. That failure taught me more than any screen I designed. Trust and transparency don’t get adopted because they’re right. They get adopted when they’re built into something people have a reason to use.