Birdie

Birdie — An AI 'operating system' for Mexican/LatAm foreign trade whose byproduct is a normalised, enriched and validated shipment corpus: it ingests data from shipping lines, port terminals, airlines, SOIA, customs brokers, ERPs, WMS and even broker email, then 'ingests, normalises and enriches' it into a single control tower covering each shipment from arrival to release. Layered on top is a document-validation product (Valid AI) that automatically cross-checks commercial invoices, bills of lading, permits, tariff classifications, declared values and weights, plus predictive delay alerting,…

Birdie offers (Alternative, Supply Chain, Geospatial, Reference) — Early-stage and thin in absolute count: at seed scale the validated shipment corpus is plausibly 10^3-10^5 shipments with their document sets. The value is per-shipment label richness (validated document pairs, terminal event timing, delay outcomes), not volume — and volume is the constraint a buyer must size first..

Three distinct buyer angles: (a) document-AI training — labelled invoice/BL/pedimento mismatch and correction pairs in Spanish-language trade documents, valuable for any customs-automation or document-extraction model; (b) Mexico port and inland-flow intelligence — terminal arrival-to-release timing and predictive-delay labels by port, an operational view absent from published port statistics; (c) customs-friction and mis-declaration indices by port, commodity or broker, usable as a border-risk indicator. Secondary: the integration map names the authoritative Mexican sources (VUCEM, SOIA, terminals, carriers) for building our own ingestion.

The data is with 3 years of history.

Coverage spans Other, US; Application Software, Marine Shipping; alternative, supply_chain, geospatial, reference; equities.