Skip to main content
Documents are part of the sub-client you create, not a separate upload step. They attach to the record they belong to — a registration document to the business’s identifying information, an identity document to the person it identifies.

​
Where they attach

Every document sits inside an identifying_information entry, in that entry’s documents array. A person’s identity document nests under that person’s entry rather than sitting loose. Both value sets — the kinds of identifying information, and the kinds of document — are listed in User values.

​
Two ways to send a file

The whole request body is capped at 10 MB, and inline base64 counts against it with about a third of overhead on top of the file’s own size. A handful of scanned documents reaches that limit quickly — send a URL instead, which is fetched separately, does not count against the body, and may be up to 30 MB.

​
Which documents

Which ones a given business needs depends on what it is and where it was formed. The sub-client tells you: missing_fields names what is still outstanding, grouped by product, and it is the list to work from rather than a table you keep on your side.

​
When one cannot be read

If Kira cannot fetch a document you supplied by URL, user.document.download.failed fires. Nothing else tells you — a document that never arrived looks the same as one never sent, so handle that event rather than assuming an upload succeeded.

RFI

What happens when something more is needed.