International Trade: Lost in Translation
We think a lot about physical/digital interfaces at Vector. We see a lot of amazing technology innovations changing the way traditional businesses operate, but as companies increasingly seek to bring processes online, the common problem remains: how do you translate the complexities of the real world into a digital one? We call it the interface problem.
Every time you see someone doing data entry, or confirming a trade, or even using a search engine to answer a trivia question, that’s really just an interface problem. Translating any unstructured format (notes, trade details, speech) into a structured digital standard (excel, reuters, google)? Interface problem.
At Vector.ai we believe that solving the interface problem is where the real benefits of AI lie. AI in the broadest sense solves interface problems by taking diverse real world inputs and learning to arrange them into understandable outputs. We think of AI as the universal travel adapter, or the Rosetta stone, of the software space.
Today, international trade has a HUGE interface problem.
The standard for international trade has historically been just an accepted procedure: agree trade, generate docs, ship, deliver docs, payment flows, repeat. That standard is beginning to change: distributed ledgers, which essentially codify and centralize the rules around legacy trading procedures, are looking to set a new digital standard.
The interface however, remains an issue.
The interface problem in international trade is currently solved by operations departments within participating companies, which is an incredibly manual burden. These departments generate, scrutinize, exchange, compare and act on documents that are all different in many small ways. In other words, they deal with and process unstructured data. Trade finance banks for instance process hundreds of millions of documents every year! That is an interface problem that should be solved with AI.
Even in a world where a distributed ledger becomes the de-facto standard for international trade, the interface between reality and ledger will still be a problem. Different document or electronic formats will still need to be standardized and brought on-ledger. Waypoints to signal good delivery will still need to be interpreted and fed into processing systems. Cargoes, and their recipients, will still need to be inspected.
This is what we’re excited about. At Vector.ai we’re using our technology as the Rosetta stone for trade processing. Today it means helping traders, banks, shippers and corporates process their trade documents, but we equally see a future where our technology provides the oracle for distributed ledger-based trade too.
The author is co-founder of Vector.ai: Solving Real-World Logistics Problems, Today.
