Everyone sells the top of the trade stack. We're building the bottom first.

Sit through enough trade-tech demos and you start to notice the trick.

The screen fills with a beautiful carbon dashboard. An AI compliance copilot answers a tariff question in plain English. A supplier risk heatmap glows green and red. It looks like the future.

Then you ask the only question that matters: “Where does the underlying shipment data come from?”

And the room goes quiet. Because the honest answer, almost every time, is: someone on your team still keys it in by hand, from a PDF.

That is the trade software industry in one sentence. Gorgeous roof. No foundation.

Trade software is a stack, and it has an order

Think of it as four layers, bottom to top.

1. Documents. What is actually on the bill of lading, the commercial invoice, the packing list, the certificate of origin. Shipper, consignee, ports, HS codes, weights, container numbers.
2. Compliance and tariffs. What duty applies, which trade agreement you qualify for, what is restricted, what the landed cost really is.
3. Supplier risk. Who you are trading with, how exposed you are, what is changing.
4. ESG and carbon. Scope 3 emissions, CBAM liability, the regulator-ready reports everyone is panicking about right now.

Almost every vendor sells you this stack top-down, because the top is where the panic and the budget are this year. CBAM is scary, so they sell a CBAM dashboard. Carbon is board-level, so they sell a carbon report generator.

We are building it bottom-up. On purpose. And we think top-down is quietly broken.

Why top-down fails

Every layer inherits the layer beneath it.

A CBAM number is only as good as the shipment data it is calculated from. A supplier risk score is only as good as the party names extracted from your documents. A landed-cost figure is only as good as the HS code and weight it started with.

So when the bottom layer is hand-keyed, and 3 to 5 percent of fields are entered wrong (which is a normal manual error rate), every shiny layer above it inherits those errors. The dashboard still looks gorgeous. It is just confidently wrong.

Garbage in. Beautiful dashboard. Garbage out.

You cannot automate the top of the stack while the bottom is still a person retyping a PDF at 11pm before a deadline. You can only make the wrongness prettier.

The unglamorous order we are actually building in

Here is our roadmap, stated plainly, including where we are not yet.

• Now: Document Intelligence. Read the documents. Return clean, structured, validated data with confidence scores, in seconds. This is the boring foundation. It is also the only layer that makes the others trustworthy. It is what we are shipping first, in early access.
• Next: Trade compliance and tariffs. Duties, FTA eligibility, restricted-party screening, landed cost. Built directly on the clean document data, so a single upload can return the extracted fields and the compliance overlay together.
• After that: Supplier risk. Scoring and monitoring built on the parties we already pulled from your documents, so you do not have to hand-build a supplier list first.
• Last: ESG and carbon, including CBAM. The layer everyone wants today, built last, because it depends on every layer below it being true.

Notice the honesty in that list. We are at the first one. Compliance, supplier risk, and carbon are roadmap, not product. Our dedicated CBAM and ESG module is planned for 2028, not next quarter.

We are telling you the order on purpose, because the order is the point.

The discipline most startups skip

Here is the part that makes investors nervous and makes the product good: we ship one module at a time, and each one has to earn real paying customers before we start the next in earnest.

No four half-finished modules. No “launch every feature to look hot for the next raise.” One layer, made genuinely good, paid for by people who use it, before we climb to the next.

It is slower. We know. It is also the only way the top of the stack ever ends up trustworthy, because by the time we get to carbon, the document layer and the supplier layer underneath it will have been in production for years.

What this means if you are the one buying

You do not have to take our word for any of this. You just have to ask vendors the quiet question.

When someone demos you an automated CBAM tool, an AI compliance assistant, or a supplier risk platform, ask: “Where does the underlying shipment data come from, and who keys it in?”

If the answer is “you upload a spreadsheet” or “your team enters it,” you are not buying automation. You are buying a dashboard that sits on top of the same manual work you already do. The foundation is the product. Everything else is paint.

We are building the foundation first, in the open, and writing down the order as we go.

Get early access to the document layer, or try it on one of your own documents and see what clean, structured data actually looks like.