Case study · Product build
A baggage-claim co-pilot, designed and shipped in 72 hours.
Airlines count on travelers missing the filing deadlines that force a payout. My Bag Claim turns those rules into a tool that tells you exactly what to file, by when — and drafts the demand letter for you.
- Built by
- Marcellus Spears
- Timeline
- 72-hour MVP
- Status
- Live · in market
The problem
The money is real. The deadlines are short. Almost nobody knows them.
When an airline loses, damages, or delays a bag, the traveler has a legitimate claim — but only if they file the right report within a legal window that can be as short as seven days. Those windows vary by whether the flight was domestic or international and by the type of claim, and they trace back to regulations most people have never read: the US DOT rules and the Montreal Convention.
Miss the window and the claim is dead. The airline is not going to remind you. That information gap is the entire reason valid claims go unpaid.
What I built
A full product, not a landing page: tracker, generator, payments, accounts.
The free deadline tracker takes the flight date, route type, and claim type and returns the exact dates the traveler has to hit, each tagged with the regulation behind it and a red / yellow / green urgency flag. The paid $19 generator produces a formatted demand letter with the right legal citations, exported as a PDF.
Behind that sit the parts that make it a real business: email accounts and password reset, a Stripe checkout with webhook handling for the one-time payment, a Postgres database tracking access and saved calculations, route protection middleware, and a user dashboard. It is deployed and live at mybagclaim.com.
The numbers
A sellable product in the time most teams spend scoping one.
72 hr
From empty repo to a shipped, paid MVP
6
Production features in the first build
$0 / $19
Free deadline tracker, paid demand-letter generator
3
Statutory deadlines computed for every claim
Under the hood
The deadline engine encodes real aviation regulation.
Each route-and-claim combination maps to its own filing windows and the specific regulation that sets them. This is the logic the free tracker runs on every calculation.
| Route | Claim type | Formal claim window | Outer limit | Regulation |
|---|---|---|---|---|
| Domestic (US) | Lost / damaged | 45 days | 1 year (DOT complaint) | 14 CFR Part 254 |
| Domestic (US) | Delayed | 21 days | 1 year (DOT complaint) | 14 CFR Part 254 |
| International | Damaged | 7-day report · 21-day claim | 2 years (legal action) | Montreal Convention Art. 31 / 35 |
| International | Lost / delayed | 21 days | 2 years (legal action) | Montreal Convention Art. 31 / 35 |
The tracker also flags urgency automatically — red under two days, yellow under seven — and never shows a traveler a negative countdown, only a clear “window passed.” Small details, but they are the difference between a tool people trust and one they close.
Honest trade-offs
What 72 hours buys — and what it deliberately leaves out.
- It is an MVP, and it is scoped like one. The build proves the core loop — calculate, generate, charge. The roadmap features (a guided claim wizard, document upload, AI-assisted document review, an expanded airline database) are intentionally not in the first release.
- The current logic is rules-based, not model-based. The deadlines are deterministic date math against regulation — which is exactly what you want for a legal window; you do not want a model guessing a filing date. AI-assisted document review is the planned next layer, added where judgment helps, not where precision is required.
- It is an informational tool, not legal advice. The product is upfront about that line, and the demand letter is a starting template the traveler owns and sends.
Why it matters for your business
The distance from “we should build that” to a live product is days, not quarters.
My Bag Claim took a messy real-world problem, encoded the domain rules into working software, wired up payments and accounts, and went live — in a weekend. That is the same speed and product instinct I bring to an internal automation: find the workflow that is costing you, build the smallest thing that fixes it, and ship it before the momentum fades.
Most automation ideas die in a backlog. The skill that matters is getting them into production, and getting the boring parts — payments, data, edge cases — right the first time.
Have a workflow that should already be automated?
The AI Readiness Assessment finds the ones worth building first — ranked by what they save you. Flat $500, refund guarantee.