A seat for capital, a home for the product.
For seven releases we built the trade itself — the supplier who makes the goods, the forwarder who moves them, the client who receives them, the operator who runs the desk. Four portals, one order, four honest views of it.
This release is three moves at once. The financier seat — the party that funds the trade — is now designed end to end. Trade OS has its own address, edma.trade, and the early-access whitelist is open. And we are introducing Growyn, the marketing engine that built the site you are reading this on.
The financier seat, designed end to end.
A real trade has one more party we had not drawn yet — the one who funds it. This release completes that seat in design: the Financier portal, fifteen screens, the full path from finding a deal to settling it.
The financier opens to their own command center. Capital deployed across active deals, fees owed and when they fall due, a payment-reputation score the rest of the network can see, and an attention queue that surfaces the two or three things that actually need a decision today — a claim raised, a repayment overdue, a milestone hit. It is the same operator-grade morning view the rest of Trade OS runs on, tuned for a funder rather than a desk.
Discovery that already knows your mandate.
From the home view the financier moves into the marketplace — every live receivable an operator has published, shown as a dense grid or a scannable list. The job here is not to browse, it is to qualify fast.
Every listing carries a fit score against the financier's mandate before they open anything — tenor, ticket size, sector, geography, and recourse preference folded into one number. A 98%-fit deal and a 60%-fit deal look different at a glance, so attention flows to the deals worth the time instead of being spent reading every card.
Disclosure is the architecture, not a toggle.
Open a listing and it appears in one of two states. Summary-only is the default: industry and sub-industry, buyer and supplier region, a receivable value band, payment terms, the financing structure, and the operator's public reputation and deal history — enough to decide whether to engage, and nothing that identifies the counterparties or exposes the economics.
The page is honest about the boundary. A visibility matrix lists, side by side, exactly what is shown now and what stays protected until disclosure is granted — operator legal name, supplier identity and factory address, the buyer entity, the exact receivable amount, order line items, the production schedule, the margin. Disclosure is requested, reviewed by the operator, and granted; it is not a switch the financier flips. The privacy holds at the data layer, so the protected fields are simply not in the financier's response until the operator releases them.
Once the operator grants disclosure, the same listing opens fully — counterparties named, the exact receivable, the order detail, the supporting documents. The financier now has everything needed to price the deal, and the operator handed it over deliberately, to someone who had already cleared the summary-level bar.
Offer, then agreement, as designed states.
From a disclosed listing the financier composes an offer — amount, rate, fees, conditions — and tracks it through to a decision, with the operator's response and any counter visible in one place. No email thread, no spreadsheet; the negotiation lives on the deal.
When an offer is accepted it becomes an agreement — and the agreement is not a single screen, it is a lifecycle drawn as explicit states. Active while the trade runs and repayments flow. Pending legal while documentation is being countersigned. Settled once it has repaid and reconciled. There is never ambiguity about where a deal stands, because the state is the screen.
And the far end of that lifecycle: a settled agreement, repaid and reconciled, with the full record of what happened preserved for the next deal with the same counterparty.
Deals you sourced yourself, in the same book.
Not every deal comes off the marketplace. A financier with their own origination — a bank that already banks the operator, a fund with a direct relationship — can add an externally-sourced private deal through a guided wizard, capturing the same structured fields a marketplace listing carries, then carry it in the same portfolio as everything else. One book, whether the deal came from the marketplace or from the financier's own desk.
And the door swings both ways. When it makes sense, the financier can promote a private deal onto the marketplace — opening it to co-investors, syndicating part of the ticket, turning a bilateral arrangement into a participation.
One portfolio, with performance and risk in view.
Every agreement — marketplace and externally-sourced alike — rolls up into one portfolio with a performance-and-risk view over the whole book: capital deployed and outstanding, yield, repayment behavior, and concentration by operator, sector, and geography. The financier reads their book the way the operator reads their desk.
Fifteen screens, one continuous path: discovery, disclosure, offer, agreement, settlement, portfolio. The financier seat is design-complete. Build is the next workstream.
edma.trade: Trade OS has a home, and the whitelist is open.
Trade OS now lives at its own address: edma.trade.
What it is. The operating system every party of a trade can sign into. Your suppliers confirm orders from their portal. Your forwarder updates shipments from theirs. Your client tracks deliveries from theirs. You see all of it on one screen, in real time — the four portals from Release #7, now with a front door the market can walk through.
The landing page leads with the product doing its job, not a stock illustration of it: a live shipment tracker on a world map, vessels in transit with ETAs and exception flags — the same operational truth a customer sees after they sign in. The pitch and the proof in one view.
Why it is a separate site. edma.app is the protocol and the token rail — verification, EDM, the Layer 2. edma.trade is the product a business logs into and pays for. Two readers, two jobs. Splitting them lets each speak plainly to the people it is for, instead of making the operations buyer read past the token and the crypto reader read past the freight.
Why you should believe it. It is not a concept site. Trade OS runs in production at EDMA Group today — $80M in trade revenue, 12 active supplier factories, and 8 destination markets moving through it. Public v1 launches Summer 2026, and the early-access whitelist is open now at edma.trade.
Introducing Growyn.
A quick introduction to a sister project, because the design is too good to keep to ourselves — and because you are already looking at its work.
Growyn is an AI marketing autopilot. Research, capture, nurture, ads, optimization, analytics — the full demand engine, run by AI with a human in the loop. EDMA is its first customer: this website, the campaigns behind the presale, the content you read — all Growyn's output. We market EDMA with the thing we are building.
The design system is finished and the surfaces are designed across the suite; implementation is in progress. Here is the structure, area by area.
The styleguide — one brain for everything.
Everything starts with the styleguide. Issue 03 is complete — fourteen sections covering tokens, type, layout, surfaces, forms, controls, feedback, navigation, data, and editorial. It is the single source of truth, so every area shares one look, motion, and voice instead of drifting screen by screen. The language is warm and editorial: Fraunces for display, Inter Tight for text, set on paper-toned surfaces with moss green and amber copper as the accents.
The dashboard — your funnel, this morning.
The dashboard is the operator's morning view of the whole funnel. Traffic, leads, opportunities, customers — the full path on one screen, with the bottleneck called out and what it is costing stated in plain numbers rather than buried in a chart. A generated daily brief reads the state of the funnel and tells you where to spend the day.
The suite, area by area.
Under the dashboard sits the engine itself — nine areas, each designed in the same language, each a real working surface rather than a settings tab.
Website. The marketing site itself — pages, a blog, SEO tooling, and an insights view — built and run from inside the product. The site you are reading this on is the reference implementation.
Convert / Capture. Forms, popups, and lead magnets, each with attribution baked in, so anonymous traffic becomes a known list and you can see which surface earned which contact.
Nurture. Email campaigns, lists, and reusable templates that carry a contact from first touch to ready-to-buy — sequences that react to behavior instead of blasting on a fixed calendar.
Paid Ads. Campaign creation, channels, audiences, and creative, wired into the same funnel everything else reports into — so paid spend is measured against the real pipeline, not a siloed ad dashboard.
Optimize. A standing loop for improving what is already running — experiments, recommended actions, and generated briefs — so the funnel keeps tightening instead of being rebuilt from scratch each quarter.
Analytics. Overview, reports, and channel- and funnel-level analysis in the same editorial language as the rest of the suite — the numbers behind the dashboard's daily brief, shown in full.
Content. A calendar, a media library, and a published view — the whole content operation in one place, from idea to live, instead of scattered across a doc, a drive, and a CMS.
CRM. A pipeline for moving real deals to closed, so the demand the rest of the engine generates lands somewhere accountable rather than leaking out the bottom of the funnel.
Funnel Plan Builder. Editable assumptions, scenario toggles, and live targets — the plan everything above executes against. Set the model, and the dashboard measures reality against it.
Design is done across the suite; implementation is in progress. It is a sharp project with a sharp design — and it is how EDMA markets itself.
What this adds up to.
Capital now has a seat at the table — designed, fifteen screens, ready to build. The product has an address the market can reach, with real volume already running through it. And the engine that will sell it is taking shape in the open.
Verify first. Then mint.
