Why we built Qootna for sovereign buyers

Most freight-tech and logistics SaaS is built for the shipper or the freight forwarder. The unit of optimisation is per-container margin, the user is a planner inside a carrier or 3PL, and the data model assumes commercial confidentiality between buyer and seller. That stack works — for them.

A sovereign buyer has a different problem. A national food agency, a strategic stockpile, a defence procurement office, a sovereign wealth-linked trader: each of them is buying for a population against a policy mandate. The questions they need answered are not "what's the cheapest container today" but:

• Can this corridor still execute under the current crisis posture?
• What's the defensible cost-per-MT we file with the budget committee?
• If we re-route via an alternate port, what does the planning surface look like for the next 90 days?
• When the auditor reviews this decision in two years, can we show the inputs?

These are research-grade questions and they need a planning surface, not a marketplace.

The trust gap

Sovereign procurement runs on defensibility. A number is only useful if you can show its provenance to a board, an auditor, or a parliamentary committee. The standard SaaS pattern — blended scores, opaque ML, "trust us" pricing — fails that test.

Qootna's engine emits every input with its source labelled. A line shows whether it came from a verified contract, a published benchmark, or a fallback. Confidence is a tier, not a vibe. When a rate is missing we tell you so explicitly, and the budget line carries the right disclaimer. Nothing renders to the user that we can't defend.

The infrastructure gap

A sovereign buyer cannot share a Postgres cluster with consumer SaaS, no matter how strong the row-level security story is. Single-tenant infrastructure is not a feature flag — it's a constraint that shapes the entire deployment topology. We host in the EU, single-tenant, with audit logging at the application layer and immutable storage for decision artefacts.

That's not a sales line — it's why we pay more for hosting and ship more slowly than we would on a multi-tenant model.

The data gap

Public sources do most of the work. UN Comtrade for national flows. WITS/UNCTAD for tariff schedules. Public AIS for vessel positions and ETAs. National port authorities for handling tariffs. Open weather APIs for the routing-condition modifier. Where carrier data improves a number, we integrate it — Maersk, DHL, Emirates SkyCargo today; more in discussion.

What we refuse to do is launder a single black-box number on top of the stack. Every result is decomposable to the inputs. If a buyer wants to challenge one of them, the trace is there.

Why this matters now

Trade is fragmenting along policy lines. Strategic commodities — grain, dairy, pharma, defence-relevant materials — are moving through more constrained chokepoints with less predictable execution. The buyer who can answer "what's our exposure and what's our alternate" in seconds is a different operator than the one who waits for a forwarder to email back.

That's the operator we're building for.