Catch the bad field before it reaches your systems.
AWBGuru validates every air waybill field against real air-freight rules — so a wrong value is caught before it ever reaches your systems.
Each check resolves to pass, warn, or fail and feeds a confidence score. You set one threshold; clean documents clear themselves and only the fields that fail reach a person.
30 free scans to start · no card required · then pay per scan or pick a plan
Validation is not OCR confidence.
An OCR score tells you how sure the model is that it read the characters correctly. It says nothing about whether the value is right. AWBGuru adds the second layer: freight-native checks that prove each field against the rules air cargo actually runs on — so a cleanly-read but wrong number still gets caught.
What runs on every air waybill.
Each is an exact, explainable rule — it resolves to pass, warn, or fail, and each result feeds the confidence score.
AWB check digit
The serial's last digit must equal the serial mod 7. Pure arithmetic — a mis-keyed or transposed number fails it outright.
Catches: transposed / mis-keyed AWB numbers.
Airline prefix
The 3-digit prefix must map to a real, known carrier — not just be three digits.
Catches: a prefix that belongs to no airline.
Route & printed city
Origin and destination must be valid IATA codes and serve the printed city name — Incheon (ICN) counts as Seoul because IATA files it under the Seoul metro code — the reconciliation a plain code lookup skips.
Catches: the “valid-but-wrong” airport misread.
HS code
The commodity code must be a real Harmonized System reference with 6, 8, or 10 digits — not merely “looks numeric.”
Catches: incomplete or invented commodity codes.
Chargeable vs gross weight
Chargeable weight can never be less than gross — chargeable is the greater of actual and volumetric weight. A stated chargeable figure below the gross weight is flagged.
Catches: a transposed or understated chargeable weight.
Currency & rate class
The currency must be a valid ISO-4217 code, and the rate class a real IATA rate-class code (M, N, Q, C, R, S, U, E, X, K, P, B, Y) — not free text that merely looks plausible.
Catches: a mis-read currency or an invalid rate-class letter.
Carrier ↔ prefix cross-match
The 3-digit prefix must agree with the operating carrier read from the flight number. Carriers hold whole prefix sets — Lufthansa's 020 (cargo) and 220 (passenger) both belong to LH — so we test membership, not a single value. A prefix owned by one airline on a flight operated by another is flagged.
Catches: a swapped or mis-keyed prefix that doesn't match the flight's carrier.
Required-field completeness
Mandatory fields for a valid waybill must be present. A blank where the format requires a value is flagged, not silently passed.
Catches: a missing origin, weight, shipper, or consignee.
You draw the line once.
Every check resolves to one of three states, and they roll up into a single document confidence score. You set one threshold that fits your risk tolerance — raise it and more fields route to review; lower it and more clear automatically. Because the correction rate drops sharply as confidence climbs, a sensible threshold sends only the genuinely hard fields to a person.
Pass
The rule is satisfied. The field clears toward delivery.
Warn
Plausible but unproven — routed to review by your threshold.
Fail
The rule is broken outright — held for a human before it ships.
Score
All results feed one confidence number that decides auto-clear vs review.
A failed field never stops the line.
When a check fails, that field routes to a review queue instead of delivering — but the batch keeps moving. Clean documents clear while someone resolves the exception, each correction feeds back to sharpen the checks, and every run leaves a structured audit receipt: the fields read, the checks run, the results, and where the record was delivered.
Validation runs on the record produced by MAWB data extraction — extraction reads the fields, validation proves them.
Air waybill validation, answered.
What's the difference between a warn and a fail?
A warn is plausible-but-unproven — it's routed to review based on your threshold. A fail breaks an exact rule outright (a check digit that doesn't compute, an airline prefix that isn't a valid three-digit code) and is always held for a human before it can deliver.
Can I tune the threshold?
Yes — you set a single confidence cutoff. Raise it to send more fields to review for tighter control; lower it to auto-clear more for throughput. It's one dial, not a rules engine to maintain.
Does a failed field stop the whole batch?
No — batches process document by document, not all-or-nothing. Only the document with the failed field waits in review; the rest of the batch delivers on its normal schedule. One bad waybill can't hold up the run.
Do I get an audit trail of what was checked?
Every run leaves a structured receipt — the extracted fields, each validation result, timestamps, and the delivery record — so you can show exactly why any document cleared or was held.
See the checks run on your own waybills.
Start with 30 free scans — no card. Watch a caught error route to review while the clean ones clear.