Here is a fact about global trade that does not make it into the glossy dashboards: a shipment's actual status usually lives in an email. Or a WhatsApp thread. Or a PDF of a customs filing that somebody scanned slightly crooked. The freight itself moves on ships and trucks, but the information about the freight moves on the same tools your aunt uses to forward you recipes. This is the unglamorous, load-bearing truth that Desteia decided to build a company around.
Desteia is a New York-based startup, founded in 2023, that calls itself "the AI platform for global trade." That phrase could describe roughly four hundred other companies, so it is worth being specific about what Desteia actually does, because the specific part is the interesting part. The company uses AI to read the unstructured mess of logistics - the emails, the messages, the documents - and pull out the facts hiding inside them. Then it does something a little unusual: it organizes those facts using graph theory.
Graph theory is the branch of mathematics that studies networks - dots (call them nodes) connected by lines (call them edges). Leonhard Euler more or less invented it in 1736 to settle an argument about whether you could walk across all seven bridges of a Prussian town without crossing any twice. (You cannot. It bothered people.) Desteia's insight is that a supply chain is, structurally, exactly this kind of object: locations are nodes, routes are edges, and a shipment is a path through the graph. Model it that way and a lot of previously invisible things - a bottleneck at a border crossing, a delay rippling three ports downstream - become things you can actually see and reason about.
There is a good line buried in that quote. The world does not have a shortage of supply-chain risk analysis; it has a shortage of risk analysis you can act on. A 40-page report telling you the U.S.-Mexico border is fragile is not helpful at 9 a.m. when you have a specific truck stuck at a specific crossing. Insight you cannot act on is just expensive trivia. Desteia's pitch is that it closes the gap between "here is a risk" and "here is what to do about your shipment."
The border as a business plan
The company could have tried to model all of global trade at once. It did the smarter thing and went narrow. Desteia focuses on the U.S.-Mexico corridor, where companies move more than $850 billion of goods a year - a number that got dramatically more interesting once "nearshoring" became the strategy of the decade and Mexico became, for the first time, the top trading partner of the United States. When a market reorganizes that fast, the incumbents' tools break first, and that break is where a startup gets built.
Desteia leans into the specificity. It integrates directly with Mexico's customs system, which gives its users the kind of port and border visibility that is genuinely hard to assemble by hand. The platform tracks shipments across ocean, ground and air, and stitches them into a single view the company likes to call a "command center" rather than a "tool." The distinction is not just marketing. A tool waits for you to open it. A command center is supposed to tell you what needs your attention before you think to ask. Building the second thing is much harder, and much more valuable, than building the first.
Who is building it
The founding team is a specific combination: Francoise Lavertu Stevens, a former Tesla executive, and Diego Solorzano and Austin Poore, both Stanford engineers. Lavertu Stevens and Solorzano run the company as co-CEOs; Poore is CTO. It is a group that looked at global logistics and concluded, correctly, that the hard problem is not moving atoms - trucking companies are quite good at moving atoms - but moving the information about the atoms. That reframing, from a trucking problem to a data problem, is the whole company.
In February 2025 the thesis got validated with money. Desteia raised an $8 million seed round led by Autotech Ventures, Nazca and Village Global, with Foundamental, Bridge Latam and Nido Ventures participating. That brought total funding to $11.5 million. The customer list already skews enterprise - major North American retailers, auto manufacturers and consumer-packaged-goods companies - and includes Elektra, one of Latin America's largest retailers. For a company of roughly thirteen people, that is a respectable set of logos to be responsible for.
The unsexy bet
What is most likable about Desteia is the part of the problem it chose to attack. It did not build a flashy visualization and call it visibility. It went to the source of the mess - the inbox, the attachment, the customs filing - and started there, with the reading and the structuring and the validating. Nobody's dream job is parsing a bill of lading. That is precisely the argument for having software do it. If Desteia is right, the supply-chain dashboard everyone has been staring at for a decade was never the product. The product is the boring machinery underneath that finally makes the dashboard true.