Ocean Booking Confirmation OCR Parser

The carrier's confirmation that space and equipment have been allocated for a shipment on a specific sailing. It carries the booking number, vessel and voyage, routing, equipment committed, empty pickup details and — most importantly — the cut-off deadlines for documentation, VGM, cargo gate-in and customs.

The documentation or SI cut-off is when complete shipping instructions must reach the carrier. The VGM cut-off is when the Verified Gross Mass declaration is due. The cargo or gate-in cut-off is when the full container must be inside the terminal. There is usually a separate customs cut-off, and on some trades an advance manifest cut-off. They fall on different days, belong to different teams, and any one of them can roll the container.

Because a date in a PDF is not a deadline anyone can see. As structured fields on the shipment record, cut-offs become visible to whoever is working the shipment and can be surfaced before they pass. The common failure is not that someone read the deadline wrong — it is that the deadline belonged to nobody.

Yes, and this is one of the more valuable cases. Amended confirmations frequently look almost identical to the original, with a changed vessel, voyage or cut-off buried in an otherwise unchanged document. Extracting both and comparing them makes the change explicit rather than leaving it to be spotted by eye.

The quantity, size and type codes committed, the empty pickup depot, the release reference and its validity period. The type code matters: a booking confirming standard 40ft equipment against a request for high cubes is a problem that surfaces at loading if it is not caught on the confirmation.

Yes, and it should be. Grouping both into one shipment record compares vessel, voyage, all four routing points, equipment and references. A discharge port or vessel that differs between the two is either an amendment somebody agreed or a change nobody noticed, and distinguishing those is exactly why the baseline is worth keeping.

Where they are stated, yes. Where they are not — which is common — the time is recorded as printed and the ambiguity is left visible rather than resolved by assumption. A cut-off stated as '18:00' on a booking involving three timezones is a genuine risk, and inventing a timezone would hide it.

Yes. Forwarder and NVOCC booking notes carry the same information in more variable layouts, frequently as formatted email rather than a structured document. They extract into the same field model, with unlabelled or ambiguous references surfaced for review rather than assigned by guesswork.

It is flagged prominently rather than recorded as an ordinary date. A booking confirmation that arrives after one of its own cut-offs has passed is not unusual — amended confirmations and late releases both produce it — and it needs immediate attention rather than filing.

Yes, and it is the right place to start. Preparing instructions from the booking rather than from a previous shipment is what prevents stale vessel names, old container numbers and superseded consignees reaching a live transport document. Reviewed booking data can be reused directly into an instruction draft.

Where the confirmation states one, yes. Transhipment materially affects transit time and risk, and a routing that gains or loses a transhipment between the booking and the Bill of Lading is a change worth noticing — particularly on lanes where the transhipment port has its own congestion or documentary requirements.

It does not predict them, but it makes them visible faster. When an amended confirmation arrives, comparing it against the stored original shows exactly what moved — vessel, voyage, ETD, or a cut-off brought forward — instead of requiring someone to read two near-identical PDFs side by side under time pressure.

Yes — the generator maps parties, routing and equipment into an SI draft you edit and download as PDF.

Dates are extracted exactly as printed and checked for plausibility. Ambiguous formats are parsed day-first, the dominant convention on ocean documents.

Screenshots parse fine — compress happens automatically and the parser reads rendered text from images.

Start with 20 documents each month on Free. Pro includes 500 documents for one operator, while Team includes a five-seat shared workspace and 2,000 pooled documents. Both paid plans are monthly and managed through Paddle.