Mishandled baggage claims settled from the bag tag: the journey, every receipt and the right limit, cited — and a person signs
Bag claims settled in two days, within the right limit, with every figure cited.
See one case, screen by screen ↓Mishandled baggage claims software takes a delayed, damaged, lost or pilfered bag claim from the passenger's claim form to a signed settlement: the bag file, booking and scan history matched, each receipt valued, the liability limit set and the reply written. Mishandled Baggage Claims adds eight agents that rebuild the bag's journey from its scans, read every receipt, set the limit from the itinerary and draft the settlement with every figure cited, for a baggage claims agent to approve and sign.
How to choose AI for airlines: a buyer's guide to agentic AI solutions →
Updated
When a bag is delayed, damaged or comes back lighter, the passenger files a mishandled baggage claim: receipts for the toothbrush and the shirt, a list of what is missing, photos of a broken wheel. A baggage claims agent then opens the bag tracing file, the scan history, the booking and the claim form in separate systems, re-keys the receipts, decides whether the US domestic $4,700 limit (14 CFR 254.4) or the Montreal Convention's 1,519 SDR applies, depreciates the contents by hand and writes a letter.
The risk sits in the judgment calls nobody has time to make the same way twice. A domestic leg on an international ticket is international carriage, so paying the domestic figure there over-pays. Depreciation and "reasonable expense" are decided agent by agent, a receipt already paid on another claim gets paid again, and a letter that says "closed R06 per Rule 28(c)" starts a DOT complaint. Meanwhile the passenger waits weeks for the price of a toothbrush.
Estimated split for a typical delayed-bag claim at a network carrier, by hand and with the solution; the 11-day baseline is the sample airline's own.
Eight agents rebuild the bag's journey, read every receipt, set the limit by itinerary and draft the settlement; the claims agent signs.
Rachel Abernathy, a Corvane Plus Gold member, flew Boston to San Francisco through Chicago O'Hare for a client meeting. Her bag, tag CV 772415, reached her hotel 31 hours late and her jacket was missing. Here is how Nora Castellanos settles the mishandled baggage claim with the agents, screen by screen, in the working solution.
Nora works in Central Baggage Services at ORD. Her Today screen shows 168 claims read overnight, 112 drafted and ready for her to sign for $27,486, 23 that need a person — 21 for her, 2 for her manager — and 33 waiting for passengers. Beside the list, the agents explain the night in plain sentences: the bag journey agent traced 41 delays to missed connections at the ORD transfer sort, and the integrity check stopped 3 claims for the fraud analyst.
Integrity check: “I stopped 3 claims for Owen: a receipt already paid in 2025, a receipt dated before the flight, and an edited repair quote. Passengers have not been told anything.”
Claim MB-26-41872 opens on bag CV 772415: CV 1418 BOS–ORD and CV 2210 ORD–SFO on Tuesday, October 13, filed October 15 at 08:12 CT. Eleven scan points run from check-in at BOS to delivery at the hotel, with the transfer sort marked as the break. The limit chip reads “Domestic · limit $4,700”, and the agents have already said what happened and what Corvane owes, every statement cited.
“CV 1418 reached ORD 26 minutes late, so the bag reached the transfer sort at 09:14, four minutes after CV 2210's bags closed.”
Clicking the break opens the sortation record: tag CV 772415 read at T3 induct 4 at 09:14:07, CV 2210's make-up lane closed at 09:10, the bag routed to the problem-bag line. The minimum bag connection at ORD is 45 minutes and this bag had 23. It was rush-tagged to CV 2216, offloaded for weight, and flew the next morning. It weighed 47.2 lb at BOS check-in and 45.2 lb at the SFO baggage office, and screening never opened it.
The bag journey agent never blames a person. It states what a scan shows, and says “no scan” where there is none.
Rachel claimed $921.00 for six items. Toiletries, clothes and a charger bought in San Francisco on October 13 come to $150.00, exactly the daily essentials cap under Rule 28.4, and a dress shirt on October 14 adds $36.00. The $95.00 hotel dinner is proposed at $0 because meals are not a bag expense. The $640.00 merino down jacket has no receipt, so it is not valued yet. The reply to Rachel is already drafted underneath, with no codes.
Proposed now $186.00 · waiting for proof $640.00 · not covered $95.00 · limit $4,700.
Nora asks why the dinner is $0. The contents and receipts agent cites Rule 28.4(b) and adds that she can still adjust it with a reason code for goodwill. She asks whether this is an international trip. The liability agent answers that all three coupons are within the US, so 14 CFR 254.4 applies, and that a ticket starting in London would put even the ORD–SFO leg under the Montreal limit of 1,519 SDR ($2,077.99).
Liability agent: “No. All three coupons — BOS–ORD, ORD–SFO and SFO–BOS — are within the US, so 14 CFR 254.4 applies: $4,700 per passenger.”
The agents ask one thing: approve $186.00 now and ask Rachel for proof of the jacket? Nora picks reason code R01, within policy, ticks that she has reviewed the cited evidence, and approves. The essentials go to her card ending 4417 the same morning, well inside the 48-hour interim payment clock. The passenger reply writer's request asks for a card statement line, an order email or a photo of the receipt, or a photo of her wearing the jacket.
“I have reviewed the cited evidence and I sign this as Nora Castellanos, Baggage Claims Agent.”
Rachel uploads her September card statement: “the jacket is the September 21 line.” The agents re-run only what it affects. The contents agent reads HALDEN OUTFITTERS, $640.00, September 21, on the card that paid the ticket; the jacket is 22 days old, so it is paid at 100% under the depreciation table. The integrity check re-runs on the new document and stays clear. The decision becomes “Sign the $826.00 settlement?”
Settlement drafter: “Settlement $826.00 = $186.00 already paid + $640.00 for the jacket. Within the $4,700 limit and your $1,500 signing limit.”
Nora signs the $826.00 settlement with reason code R01 and sends Rachel the reply in plain English. The audit trail for MB-26-41872 shows every agent step and both of her decisions, each entry chained to the one before; Verify the chain recomputes 21 entries, unbroken. The $40.00 checked-bag fee had already been refunded on October 14 under the DOT 12-hour rule, and the evidence pack exports in one click.
“Every step an agent took and every decision a person made, in order, each entry chained to the one before. Nothing can be edited or removed.”
The batch holds 112 claims for $27,486.40, each under $500, clear of the integrity check and every figure cited: 86 delayed-bag essentials, 17 damaged bags on a partner repair quote and 9 small missing contents with proof of purchase. Nora samples three the agents chose across the groups — including Henry Albright's $145.00 wheel repair, lower than the $214.00 depreciated replacement — unticks any she wants out, and signs the rest with one reason code.
Claims coordinator: “Each claim is under the $500 batch ceiling, every receipt is matched to an item under Rule 28.4 or 28.6, the integrity check cleared it on its own, and every agent was at least 90% sure.”
Diane Whitcomb's $612.40 receipt from Northgate Apparel matches one paid on MB-25-30914 in June 2025 to Craig Whitcomb at the same address: same store, items, prices, total and minute of the day — only the date differs — and the image is 97% similar once the date is masked. Claims agents see the evidence but cannot clear the stop; Owen Brandt decides. The passenger is never told “fraud”.
Asked whether the match could be innocent: “Possibly … a family member may have re-sent an old receipt by mistake. Asking her for the card statement line would settle it.”
The same mishandled bag, seen by the five people who carry it — what their week looked like, and what it looks like now.
Eight agents rebuild the bag's journey, read every receipt, set the limit by itinerary and draft the settlement; the claims agent signs.
Runs each claim through the other agents in order, watches the service clocks and assembles the proposal a person signs; starts the integrity check in parallel and routes to a person when any agent is unsure or a figure is above the signer's limit.
Reads the claim form, receipts and photos and matches them to the bag file, booking, ticket and loyalty profile, citing the page and region of each value.
Rebuilds the bag's journey from scan messages, sortation logs, load records, rush tags, weights and courier proof, compares the transfer time with the minimum bag connection, and states the cause in one sentence.
Matches each receipt to a declared item and values it under the expense and depreciation rules: the daily essentials cap away from home, items that are not bag expenses, and depreciation by age.
Decides the regime from every coupon on the ticket — 14 CFR 254.4 when all points are in the US, the Montreal Convention at 1,519 SDR at the day's rate otherwise — and applies excess valuation, the mobility-device rule, notice windows and the last-carrier rule.
Works on its own: fingerprints receipts against 24 months of claims, tests dates against the bag's delay window, and finds edited images and repeat claimants, with the evidence side by side for the fraud analyst.
Writes the settlement from the cited findings only, with every figure traced and the payment method the passenger chose.
Writes replies and requests for proof in plain words, in the passenger's language, asking for one thing with acceptable alternatives.
Signs every payment and reply; the manager above $1,500 and any goodwill. The agents propose; a named person decides.
Every claim has a conversation beside it. Ask in plain words and the agent that did the work answers, citing the scan, receipt, coupon or rule. Nothing changes until you approve.
Why was the bag late?
CV 1418 arrived at B9 at 08:51, 26 minutes late. The bag was read at the T3 transfer sort at 09:14:07; CV 2210's make-up lane closed at 09:10. The rush flight CV 2216 offloaded 38 bags for weight, so it flew the next morning.
Why is the dinner $0?
Rule 28.4(b): meals, hotels and transport not caused by the bag's delay are not covered for a delayed bag. Rachel's flights were not disrupted, so the disruption-care rule does not apply. You can still adjust it with a reason code if you want to pay goodwill.
What if the jacket were 8 months old?
Clothing 6–12 months old is paid at 80%: $512.00 instead of $640.00. Her statement shows Sep 21, 2026, so it is 22 days old and paid in full.
Is this an international trip?
No. All three coupons — BOS–ORD, ORD–SFO and SFO–BOS — are within the US, so 14 CFR 254.4 applies: $4,700 per passenger. If the ticket had started in London, the Montreal limit of 1,519 SDR ($2,077.99) would apply even to the ORD–SFO leg.
13 screens from the working solution, on its sample data. Pick one to see it large.
The thresholds and clocks every claim is checked against are settings. Each change applies to new claims from the moment it is saved, and is logged.
| Setting | Default | Choose from |
|---|---|---|
| Essentials cap per day | $150 | Rule 28.4 · while the passenger is away from home |
| Claims agent signing limit | $1,500 | Above this the manager signs |
| Batch signing ceiling | $500 | Per claim, integrity clear only |
| Interim expenses paid within | 48 hours | Service standard CBS-3 |
| Decision within | 5 business days | From complete proof |
| Integrity look-back | 24 months | Changed Oct 14 by Sam Kowalski, from 12 months; logged |
| Reminder to passenger | Day 3 and day 7 | Then the claim waits; it never closes by itself |
| Agent confidence for a person to review | Under 90% | Below this the claim always goes to a person |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (SITA Baggage IT Insights 2026 via CAPA — 24 million bags mishandled in 2025, $6.3 billion, about $260 a bag, transfers 39%). Regulations named: 14 CFR 254.4 ($4,700 per passenger, domestic); Montreal Convention Arts. 17, 22(2), 31 and 35, with the limit revised to 1,519 SDR; 14 CFR 382.131 (mobility devices); the DOT checked-bag fee refund rule (14 CFR Part 260); 14 CFR Part 234 (DOT reporting). Rule numbers, reason codes, clocks and thresholds shown are the sample airline's own. People, the airline, flights, shops and claims named on this page are fictional — characters and sample data in the working solution.
Mishandled baggage claims software takes a delayed, damaged, lost or pilfered bag claim from the passenger's claim form to a signed settlement: the bag file, booking and scan history matched, each receipt valued, the liability limit set and the reply written. Mishandled Baggage Claims adds eight agents that rebuild the bag's journey from its scans, read every receipt, set the limit from the itinerary and draft the settlement with every figure cited, for a baggage claims agent to approve and sign.
For a trip with every point in the United States, 14 CFR 254.4 says a carrier may not limit its liability below $4,700 per passenger. For an international journey the Montreal Convention limit is 1,519 SDR, which the solution converts at the day's rate ($2,077.99 in the demo). On domestic trips, wheelchairs and other mobility devices fall outside that limit under 14 CFR 382.131 and are paid on their original purchase price; the sample airline applies the same rule internationally by its own policy. Excess valuation bought at check-in raises the limit — in the demo, $2,000 of excess valuation takes a domestic claim's limit to $6,700.
When the ticketed journey has any international point. A domestic leg that is part of one international journey on one ticket is international carriage, so the Montreal limit applies to it too. The liability agent reads every coupon on the ticket and cites the rule and the itinerary for the limit it applies.
Under the Montreal Convention, written notice is due within 7 days of receipt for a damaged bag and within 21 days after delivery for a delayed bag (Art. 31). If a bag has not arrived 21 days after it should have, the passenger can claim it as lost (Art. 17), and the right to damages ends if no court action is brought within 2 years (Art. 35). The liability agent flags late notice with the dates, and the person decides — a manager can approve goodwill.
An independent integrity check fingerprints every receipt against 24 months of claims, tests receipt dates against the bag's actual delay window, and looks for edited images and repeat claimants. It never reads the drafter's work and can only stop a claim, never refuse it. A named fraud analyst sees the evidence side by side and clears, holds or refers it, and the passenger is never told “fraud”.
In the working solution, essentials bought while away from home are paid up to $150 a day with receipts, nothing is paid for days spent at home, and meals, hotels and transport are not bag expenses. Interim expenses are paid within 48 hours while missing contents wait for proof, and contents are valued by age under the depreciation table. Each of these is a setting you configure to your conditions of carriage.
The claim shows when the checked-bag fee has already been refunded under the DOT 12-hour rule, so the settlement does not pay it twice and the reply says so. Each month the claims coordinator drafts the DOT mishandled baggage report under 14 CFR Part 234 from the bag journeys and claims, checked against the tracing files, for the manager to review and e-sign.
Yes. The agents propose and never pay, refuse or send anything. The baggage claims agent approves, adjusts, asks for proof or declines each item and signs with a reason code up to $1,500; the Central Baggage Services Manager signs above that and any goodwill; the fraud analyst decides every integrity stop. Every step is on a chained audit trail.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your conditions of carriage, signing limits, scan and tracing feeds and a sample of past bag claims — and it goes live once every quality gate has passed. In the working solution, setup ends with 200 past claims run side by side before go-live at the first station.
Turns leave on time: a slipping task is caught early, with a fix the controller approves.
1-min demoGround Handling Invoice ReconciliationEvery handler line tied to its turn, its rate and its approval before a dollar is paid.
1-min demoGround Handling SLA ManagementPenalties collected, not waived: every SLA line backed by the flights behind it.
1-min demoWe'll run Mishandled Baggage Claims on a sample of your own past bag claims, scan messages and conditions of carriage.