It learns each format once
The first statement from your bank teaches it the layout. Every one after that is read instantly, with no re-learning.
Data entry is how books fall behind
Every agency has the folder: statements, payroll exports, invoices, timesheets, all waiting for someone to key them in. So the books run weeks behind reality, and every report built on them is a report about the past.
Intake takes the folder instead of the typing. The first statement from your bank teaches it the layout. Every one after that is read instantly, with no re-learning.
And because it is reading financial data, it is built to distrust itself: rows that do not add up become questions, not entries.
How it works
1
Drop the file in
Bank statements, payroll sheets, invoices or timesheets, as CSV, TSV, TXT and PDF.
2
It reads and checks
Opening balance plus transactions must equal closing balance. Anything that does not add up, or looks ambiguous, becomes a question.
3
Answer the queue
One question at a time, with the source row shown, so you decide the edge cases and it handles the rest.
4
Preview, then import
A dry run shows exactly what would be written before anything is. After import, every document can be undone on its own.
Built to be distrusted
It checks the arithmetic
Opening balance plus transactions must equal closing balance. Rows that don't add up become questions, not entries.
It asks instead of guessing
Anything uncertain goes to a review queue, one question at a time, with the source row shown. It never guesses on money.
It matches across documents
Payroll rows are paired with the bank debits that paid them, with the evidence shown and a confidence score on each match.
Nothing is written silently
Reading never touches your books. You preview every import as a dry run first, and every import can be undone per document.
It is never compulsory
The manual path always exists. If a document confuses it, or you just don't trust it yet, type the entry in and carry on.
What you can feed it
The exports your bank, payroll and payment platforms already emit. No migration project, no new format to produce.
Bank statementsPayroll sheetsInvoicesTimesheetsCSV, TSV, TXT and PDF
One honest limit: scanned PDFs are refused with an explanation, not guessed at. If the file is a photo of a statement rather than text, Intake says so instead of inventing numbers.
Where it sits
Intake is part of the Money department of the Agency OS. Included from the Studio plan. One credit allowance covers every module, so there is no per-agent pricing and no surprise line items.
Compare plansWorks with
Questions owners ask
What file types does it read?
CSV, TSV, TXT and PDF. Scanned PDFs are refused with an explanation rather than guessed at: if the file is an image of a document instead of text, Intake tells you so instead of inventing numbers.
What happens when it is not sure about a row?
It asks. Anything uncertain goes to a review queue, one question at a time, with the source row shown next to the question. It never guesses on money.
Can it break my books?
No import happens silently. You preview every import as a dry run first, and every import can be undone per document. Reading a file never touches your books at all.
Do I have to use it?
No. The manual path always exists. If a document confuses it, or you just do not trust it yet, type the entry in and carry on.
Retire the retyping
Join the waitlist and Intake is included from the Studio plan when the OS opens. One email, no card.
Get early access
Free plan at launch. No card, and one email, when it opens.