Skip to content
Platform

Three engines.
One architecture.

A typed, versioned pipeline turns disconnected supply-chain signal into a planning surface a sovereign buyer can act on.

analysis · live
QTN-2042
PAK → QATRice (long grain) · SEA

Karachi to Hamad Port via Gulf of Oman: 8.2 days median transit, 2.1× CO₂ vs. regional sourcing, $47/MT delta under current fuel pricing.

ENV
412 kg/MT
CO₂ equivalent
SPEED
8.2 days
Median · P90 9.4
COST
$427/MT
Total landed
Routing through 47 AIS waypoints…
Weather modifier: +0.6% · monsoon residual
Spoilage risk: 0.4% · ambient hold
3
Engines per query
12
Engine outputs typed
0+
Data sources orchestrated
Free
Tier-1 carrier integrations

Architecture

From raw signal to defensible decision.

Sources stream in, ingestion normalizes and enriches, the three engines run in parallel, outputs assemble into a narrative, a chart, a map, and an exportable report.

01

Sources

Live signal from carriers, vessels, weather, and public trade datasets.

  • Carrier APIs · Maersk / DHL / Emirates
  • AIS vessel positions
  • Weather · NOAA + ECMWF
  • UN Comtrade · national priors
02

Ingestion

Typed, versioned normalization. Every row carries method + source + timestamp.

  • Normalize schemas
  • Validate & dedupe
  • Enrich with priors
  • Version every row
03

Engines

Three models run in parallel against the same query. No black box.

  • Environmental · CO₂ / H₂O / waste
  • Speed · median / P90 / crisis
  • Cost & Waste · landed + spoilage
04

Outputs

Assembled into an analyst surface the principal can read in the elevator.

  • Narrative · 2–4 sentences
  • Map · route + waypoints
  • Charts · engine deltas
  • Report · exportable

The engines

Three measurements. One vector of evidence.

ENV · 01

Environmental.

Quantifies CO₂, water, and waste per route.

Combines modality CO₂ coefficients, distance from real coordinates, and product factors (cold chain, packaging, density). A live weather modifier folds in routing condition variance — visible to the user, never opaque.

Inputs

  • · Product
  • · Origin / destination
  • · Modality
  • · Cold-chain profile
  • · Active scenario

Outputs

  • · co2KgPerMt
  • · waterM3PerMt
  • · wasteFactor
  • · scope3Breakdown
SPD · 02

Speed.

Turns transit time into a planning surface.

Median + P90 + risk-factor decomposition over real route history, with crisis scenarios pre-modeled. Toggle a Red Sea closure or a Hormuz incident and see the delta against current routing in seconds.

Inputs

  • · Product
  • · Origin / destination
  • · Modality
  • · Crisis scenario toggle

Outputs

  • · medianDays
  • · p90Days
  • · riskFactors[]
  • · crisisDelta
CST · 03

Cost & Waste.

Where shipping cost meets perishability.

Base cost × fuel adjustment × spoilage rate × waste factor. The number you actually plan against, not a generic per-MT estimate from a logistics white paper.

Inputs

  • · Product
  • · Origin / destination
  • · Modality
  • · Market context

Outputs

  • · baseCostUsdPerMt
  • · fuelAdjustment
  • · spoilageRiskPct
  • · totalLandedCost

Cost & Waste · deep-dive

Why transit time becomes landed cost.

Grain loses 2.6% over 30 days. Leafy greens lose 42%. Same route, different products, very different landed cost. The engine models this per commodity so the number on the brief is the number you pay.

Spoilage risk curve

Perishability compounds with transit.

The Cost & Waste engine models risk per commodity. Same route, different products, different reality.

0%10%20%30%40%5d10d15d20d25d30dTRANSIT DAYSLeafy greensTomatoesGrain (bulk)

At 30 days

Leafy greens

42%

Landed cost delta +$118/MT

At 30 days

Tomatoes

35%

Landed cost delta +$74/MT

At 30 days

Grain (bulk)

2.6%

Landed cost delta +$4/MT

Output

Three views. One analyst surface.

Every analysis assembles into a tabbed surface — landed cost, resilience scoring, sourcing matrix — each with the engine evidence folded in and a two-sentence narrative a principal can brief from.

  • Every number links to its source and method.
  • Every view exports — PDF, CSV, or API.
  • Every analysis is versioned, so briefs stay reproducible.

Query · QTN-2042

PAK → QAT · Rice · SEA · Karachi / Hamad

45 MT · Q3 2026
Total landed$427/MT
Base shipping
$312
Fuel adjustment
+$42
Spoilage premium
+$31
Port & handling
+$42

At current bunker pricing this route is $47/MT above 12-month median. Fuel explains 60% of the delta; the remainder is monsoon routing residual.

Data sources

Carriers, ports, and public signal.

Status reflects current commercial integration. Honesty is non-negotiable.

Maersk

Container schedules + carrier metadata

Integrated

DHL

Air freight visibility

Integrated

Emirates SkyCargo

Cold-chain SLA + transit times

Integrated

Public AIS

Vessel positions + ETA inference

Integrated

Open weather APIs

Routing condition modifier

Integrated

UN Comtrade

National trade priors + volumes

Integrated

Hamad Port

Local berth & dwell signal

In discussion

Qatar Airways Cargo

Outbound air-cargo pricing

In discussion

Security posture

Government-grade by default.

Encryption

TLS 1.3 in flight, AES-256 at rest.

Audit trails

Every admin and analyst action logged immutably.

Access control

Invite-only. Three roles. Cookie-bound sessions, no third-party identity broker.

Hosting

EU-hosted, single-tenant Postgres, no shared compute with consumer SaaS.

No SOC 2 claim until we hold the certificate.

Ready when you are

See it on your own routes.

Invite-only access. We reply within one business day.