Ask a question in plain English. You get a plain-English summary and the contract's own words underneath it — quoted, with the section and page. We all end up on the same page, and every answer a pilot corrects makes it better for the next guy.
Enter Invite CodeInvite only during beta
Yes, but not unilaterally — it takes unforeseen operational circumstances and your consent, and further limits still apply. The exact language and limits are below — read it for yourself.
Section 12.D.1, p. 176
Ask around the crew room about a pay question and you'll get a few confident answers that don't agree. The contract has the actual language, but it's hundreds of pages and nobody's digging through it on a layover. So the story that gets repeated becomes the answer. I built this so we can look it up instead.
I didn't build it to start fights. Most of the arguing I've seen comes from people being wrong about what the contract says, and that goes both directions. It just quotes the book. Sometimes the answer isn't the one you were hoping for.
It finds the language and shows it to you. What you do with it is your call, and your rep's.
Section and page on every answer, quoted word for word, so you can open your own contract and check.
More pilots asking and flagging means more of the wrong answers get found and fixed. That part only works if we do it together.
Three steps, and the last one is the important one.
Type like you're texting a buddy on the line. It understands pilot shorthand — ADG, FDP, WOCL, deadhead, all of it.
You get a plain-English summary backed by the actual contract language — quoted word-for-word, never paraphrased into something it doesn't say.
Every answer cites the section and page so you can confirm it in your own contract in 30 seconds. You're never asked to just trust it.
Real questions from the log, answered with the language and the page.
It will be wrong sometimes. It is built to not hide that. Every answer is labeled and every answer shows where it came from.
Every answer cites the section and page, and quotes the contract word-for-word. You verify it yourself — the tool never asks for blind trust.
Ask the way you'd ask a rep. It maps your everyday wording to the right clauses across the whole agreement.
If the contract clearly answers your question, it says CLEAR. If the answer depends on your situation, it says DEPENDS. If two sections contradict each other, it says CONFLICT. If it can't find it, it says NOT FOUND. It never pretends to know more than the contract says.
Every answer has a thumbs down. One tap when something reads wrong, and a box if you want to say why. A line or two about what it missed helps a lot, but the tap alone is enough.
Flagged answers come to me. I can fix what the tool finds and how it's labeled. I can't change what the contract says — that's quoted straight from the PDF with the page on it, so you can always check me.
Some questions don't have a clean answer. The contract is vague, or two sections pull different ways. When that happens the tool says so instead of picking a side. That's not it failing. That's the honest answer.
Whatever gets caught is caught for everybody. The next pilot asking that question gets the fixed answer instead of the same bad one.
This runs on a large language model. They're fast and they're useful, and they also make things up. Not once in a while — it's how they work. It'll hand you an answer in a clean sentence with a page number on it, sound completely sure, and be wrong.
I couldn't get rid of that, so I built around it. The tool doesn't answer from memory. It pulls the actual contract language and quotes it, and puts the section and page right there so you can open your own book and check it in thirty seconds. If it can't find something, it says so instead of filling in the blank.
That catches most of it. It doesn't catch all of it.
The rest is us. When something reads wrong, hit thumbs down and a pilot who flies that contract takes a look. Every one caught is caught for everybody — the next guy asking that question gets the right answer instead of the same bad one. I'm one pilot asking questions on my days off. I'm not going to find them all.
That's really the whole thing. The machine is fast. We know the job. Neither one is enough on its own.
You should be able to look up your own contract without wondering who sees it.
Your questions are not sold, rented, licensed, or shared with any third party. There is no advertiser, no data broker, and no analytics partner receiving what you ask.
This tool is independent. No airline and no union has access to what you ask, and none of them are given reports, summaries, or counts about you.
Questions are reviewed to find where the tool gave a bad answer, so it can be fixed. That's the whole purpose.
Sick leave, fatigue calls, a pay dispute, whether they can do what they just did to your schedule. Those are exactly the questions this is for.
Over 20 years flying the line. I got tired of digging through hundreds of pages to answer a simple question. So I built a better way.
Independent and pilot-funded. Just your contract, made searchable and cited — with no agenda about how you use it.
This finds and shows you contract language. It is not a lawyer and does not give legal advice. Always verify important decisions with your union rep or MEC.
Answers are generated by AI from your actual contract. AI can make mistakes — which is exactly why every answer is cited, so you can check the source yourself before relying on it.
Both are invite only right now. Your airline not listed? Let us know.
Beta is invite only while we find out what it gets wrong.
Enter Invite CodeInvite only during beta. Got a code from another pilot? That's all you need.