CloseoutReady is sold and operated by AERA. Effective 10 November 2026.
Tax status: Exempt dealer (osek patur) · Business number: 315971192 · Contact: aeraops@gmail.com
Sorting runs on our own servers
CloseoutReady decides where each document goes in the manual with its own rules: it reads the text of each file, matches it against CSI divisions, document types and your corrections, and holds anything it is unsure of in the Inbox for you. Text in ordinary PDFs and Word files is read on our servers, and scanned pages first go through text recognition (OCR) that also runs on our servers. None of this sends your files to an outside AI service.
Outside AI that reads scanned pages (only if you allow it)
The account owner chooses, in Settings, whether scans may be read in the cloud. Nothing is pre-selected and nothing is sent until the owner chooses "Allow cloud scan reading". With that choice, when our own OCR cannot read a scanned file or photo well, images of up to its first two pages are sent to Google Gemini (Google's Gemini API), which returns the text it can see. It is asked only to copy the printed words; it does not decide where the document goes, and its answer is checked by the same rules as any other text. Google receives the page image only to return that text. With "Keep scans on our servers only", no page leaves our servers for an AI service. The choice can be changed at any time, and each page re-checks it before it is sent.
On 18 November 2026 we corrected the words of this choice: they used to name a different service and say that ticking it turned nothing on. Every earlier tick was set aside and owners are asked again with these words.
Outside AI that reads files that need help (only if you allow it)
This is a separate choice from reading scanned pages. The account owner chooses it when setting up the workspace or in Settings; nothing is pre-selected and nothing is sent until the owner allows Google's Gemini. With that choice, when our rules hold a file because they cannot place it with certainty, the text we extracted from that file (at most its first 12,000 and last 2,000 characters) and its name are sent to Google's Gemini through Google's business API, which suggests a CSI division and quotes the lines that show it.
We check every quote against the document. The file is filed only when the AI's reading is backed by those quotes and agrees with what our own rules found; otherwise it stays in the Inbox with the AI's reading shown as a suggestion, and you decide. Files our rules already placed are never sent. Text that looks written to instruct software is not sent at all and is flagged for you. We keep the AI's answer with your company, so the same file is not sent again.
Each paid project includes 300 AI reads; each account gets 20 to try on its first unpaid project. More AI reads for one project: 250 reads for $19 or 1,000 reads for $59. One AI read is one file. Reading the same file again costs nothing unless the project's divisions or packages change, and a read that fails is given back. When a project has no AI reads left, or the owner has not allowed it, files are sorted by our rules and your review, exactly as without AI.
Cloud vision (off)
The app also contains an optional document classifier that would send the first page of a hard-to-sort file, as an image, and its file name to Anthropic. It is not used: the scan-reading choice above does not cover it, and it would need its own, separate consent.
What we do not do
- We do not use your documents to train AI models, ours or anyone else's.
- We do not send your files to consumer chat apps; providers are reached through their business APIs.
- Covers, tabs, bookmarks and the table of contents are built from templates and your project details, not written by AI.
You stay in charge
Automated sorting can be wrong. You review the manual, move anything misplaced and approve the final export; the Terms explain that the manual you export is the one you approved. The full list of outside services is on the sub-processor page. Questions: aeraops@gmail.com.

