Ask Tabrillo to read what is between the two of you. It goes through the lines one at a time, says which ones a reasonable person might read differently, and gives you the figure both ways — with it, and without it. Then the two of you decide.
The bill is right and it still feels wrong. Somebody did not drink. Somebody was not there for the Tuesday. The deposit came back and nobody remembers who put it down. Everyone can see the total and nobody wants to be the person who brings it up.
Two things, and most apps do neither. The first is being able to see where a number came from — which expense, whose share, how it was split, who changed it since. The second is being able to ask whether it was fair, which is a different question with a different answer.
The first one is the receipt · the second is below.
The figure with a line and the figure without it are both computed from the same entries the receipt shows. The model reads and writes the sentences; it never touches a number, and it has no way to change one. That is why a reading can be checked rather than believed.
These are the lines it looks for: a round somebody did not touch, a dinner one of you missed, a deposit that came back, a cost that was really one person’s. It names them, says why somebody might read each one differently, and shows what leaving it out would make the balance.
It proposes and never rules. Nothing moves on its own — if you both agree, one of you edits the expense, and the edit is in the history like every other.
It says what it would leave out, and what that would make the balance.
A free account gets one reading a month, which is enough to see what it does. Pro raises it to ten. The receipt underneath it — every expense, every share, every edit — is free at every tier and always will be, including without an account.
It does not, and it is built so that it cannot. It has no way to move money, no way to change an expense, and no vote. What it has is the ledger and a willingness to say out loud what everybody is already thinking — which is usually all an argument needs.
Open the balance between the two of you and ask for a reading. Tabrillo goes through the lines, names the ones a reasonable person might read differently — a round somebody did not drink, a dinner one of you missed, a deposit that came back — and gives the figure with each one and without it. It proposes; the two of you decide.
That is the case this is for. A balance can be correct and still feel wrong, because fairness is about which costs belonged to whom rather than about arithmetic. The reading separates the two: the receipt shows where the number came from, and the reading says which lines are arguable.
No. It proposes and never rules, and nothing changes on its own — it cannot move money or edit an expense. If you both agree, one of you makes the change, and it appears in the history like any other edit.
The model writes the sentences and never the figures. Both amounts — with the disputed line and without it — are computed from the same entries the receipt shows, so anything it says can be checked against the ledger rather than taken on trust.