Laytime starts with the Statement of Facts.
Proteus extracts every event from your SoFs automatically and feeds clean numbers into your laytime workflow. Source-traced. Audit-ready. In seconds, not hours.
The clock runs on your data. Make sure your data is right.
Laytime is a countdown. Every minute counts either against you as demurrage or for you as despatch. The math is exact.
What isn't exact is the data. Timestamps scraped from scanned PDFs, retyped into Excel, cross-checked against charter party terms by hand. One bad timestamp and your clock shows 47 hours of demurrage when it's really 43. One missed event and despatch you earned never makes it to the claim.
If the input is wrong, the output is wrong — no matter how good your laytime tool is.
The input to your laytime calculation is a pile of Statements of Facts.
Before any laytime math, there's a data-entry step. Somebody reads the SoF and types events into your laytime tool. That's where the errors enter the clock — mis-read times, missed events, duplicate entries, Chinese ports transliterated three different ways across three different voyages.
Proteus does the reading. You keep the calculation workflow you already have.
Upload any SoF — typed, scanned, handwritten, bilingual — and Proteus returns structured events with source citations. Export to Excel, CSV, or JSON and feed straight into your existing laytime tool. The clock starts running against clean data.
If you want the full calculation too, Proteus includes an auto-demurrage calculator that runs on top of the extracted events. Optional. Use it or don't.
The SoFs that break your laytime clock are the ones Proteus is built for.
When a bilingual SoF or a handwritten stamp-layered scan throws off your timestamps, the whole calculation is wrong. Proteus reads them correctly the first time.
Typed tables
From well-run modern ports with structured SoF tables.
Scanned & stamped documents
From Paradip, Kandla, and other ports where the SoF is still pen-and-ink with overlapping signatures and mixed date formats.
Bilingual Chinese/English
From Zhangjiagang, CJK, Lianyungang, and other terminals that issue tables in both languages.
HTML-rendered agent SoFs
From Morocco and other digitized terminals where the SoF is delivered as a structured web export.
Handwritten & mixed date formats
DD/MM/YYYY, MM/DD/YYYY, 09-July-2024, 0800 LT, 10H10. Proteus normalizes them all.
Commingled pages
SoF tables buried inside scans with cover pages, Master's remarks, cargo manifests, and correspondence. Proteus isolates the SoF pages automatically.
Multi-tender NORs
Notices tendered two, three, or four times after rejection or re-inspection. Every tender captured in sequence.
Letters of protest & disputed remarks
Master's and receiver's remarks that contradict each other on the same document, captured side by side for review.
If your analyst can read it, Proteus can extract it.
Every event Proteus extracts links back to the source SoF.
When your laytime calculation hits a dispute, you show the source — not the screenshot. Every extracted event points back to the exact line in the PDF, so your downstream calculation inherits that provenance.
No black-box extraction. No “trust the AI.” Just faster typing, with the source one click away.
Try Proteus on one SoF.
Pick a real Statement of Facts from a recent voyage. Upload it. See what Proteus extracts. If the accuracy meets your standard, you'll know in minutes whether this feeds into your laytime workflow cleanly.
Free. No commitment. Account creation happens after you've seen it work.
Frequently asked questions
Does Proteus do the full laytime calculation?
What laytime tools does Proteus output plug into?
How does Proteus handle reversible vs. non-reversible laytime?
What about NOR tender and commencement validation?
Can Proteus handle proration for part cargoes?
Is my data confidential?
What does it cost?
Your laytime clock is only as accurate as your data.
Built by Greywing.