Why we created Marintel.

Every maritime S&P transaction involves hundreds of documents. Contracts, certificates, survey reports, correspondence, legal opinions, bank conditions. These documents arrive from different parties, in different formats, at different times. They contain contradictions. They reference each other in ways that aren't obvious. They have deadlines embedded in clause language that nobody has time to read carefully in a 48-hour due diligence window.

The problem isn't that the deals are bad. Most maritime transactions are commercially sound before due diligence starts. The problem is that the process for managing the documents — and finding the risks buried in them — hasn't materially changed in decades. Email threads. Shared drives. PDF annotations. Junior lawyers billing hourly to check delivery dates against payment schedules.

We built Marintel to change that. Not with a generic document storage tool — there are plenty of those — but with a platform that actually understands what maritime transaction documents say, how they relate to each other, and what the discrepancies between them mean for the deal.

The specific problem we set out to solve

The catalyst for Marintel was watching a maritime deal come close to collapse because of a contradiction that was visible in the documents from the start. The refund guarantee issued by the yard's bank used different payment milestone definitions than the building contract. The mismatch only became visible when a payment was due and the buyer tried to call the guarantee — and found the coverage didn't match what they thought they'd paid for.

The documents were in the data room. Both of them. The contradiction was there if you read them carefully side by side. But in a deal involving hundreds of documents and a team working under time pressure, careful side-by-side reading doesn't happen reliably. It takes exactly the kind of sustained attention that document volume makes impossible.

That's the problem we built Marintel to solve: not document storage, but document intelligence. A system that reads the whole deal, understands the structure, and surfaces the contradictions that humans miss not because they're careless but because there are simply too many pages.

Why maritime specifically

Maritime transactions have specific structural characteristics that make a generic document tool insufficient. The contract forms are standardised — NSF Saleform, Nipponsale, SAJ, NEWBUILDCON — but the risks aren't always where generic legal models look for them. The deadline that matters might be a class survey expiry embedded in the delivery clause. The contradiction that matters might be between a payment milestone and a refund guarantee period.

Maritime professionals know these risks. They've seen these patterns. But knowing the patterns and being able to check every document for every pattern, across a 200-document data room, in 48 hours, are different things.

Marintel carries the pattern knowledge. The AI pipeline carries the clause libraries built for maritime contract forms. The combination — domain expertise encoded into an always-on, full-document reading system — is what makes Marintel different from a general-purpose document AI pointed at shipping contracts.

What we're building toward

We started with the document intelligence problem because it's the most acute pain in maritime transactions. But the larger goal is to build the infrastructure that makes maritime deals faster, safer, and more transparent for every party involved — buyers, sellers, brokers, lawyers, banks, and class societies.

A deal room that every party trusts because the record is immutable. A closing archive that survives warranty periods. A document intelligence layer that finds the risks before they become disputes.

That's what we're building at Marintel.