How automated ingestion works

Automated ingestion lets DASH watch a source location — such as a folder, mailbox or vault — and turn every new document it finds into a certificate, without anyone uploading anything by hand. You configure a source channel on a material, DASH polls it on a schedule, and each new document becomes a certificate waiting for your approval.

The ingestion flow

Once a source channel is active, DASH repeats the same cycle roughly once a minute for every channel that is due:

flowchart TD
    A[Poll source channel] --> B{New document found?}
    B -- No --> A
    B -- Yes --> C{Already ingested?}
    C -- Yes, duplicate --> D[Skip document]
    C -- No --> E[Create certificate]
    E --> F[Ingest and map the data]
    F --> G[Certificate reaches Awaiting approval]
    D --> A
    G --> A

Poll — A dispatcher checks each active channel about once a minute and starts a poll for any channel whose next poll is due.
Discover — DASH lists the documents at the source and picks up anything new since the last successful poll.
De-duplicate — Each document is checked against everything already ingested. Matching is by the source’s own document identifier and, as a backstop, by a hash of the document content — so a file that has been renamed or re-uploaded is still recognised and skipped.
Create — For each genuinely new document, DASH creates a certificate.
Ingest — The document is read and mapped using the channel’s pinned definition map, exactly as if you had uploaded it yourself.
Await approval — The certificate lands in Awaiting approval, ready for you to review, sign and submit.

Where the certificate goes next

Auto-created certificates enter the same lifecycle as ones you create by hand — they simply skip the draft and upload steps. From Awaiting approval you review the mapped data, sign, and submit. See Review auto-created certificates.

What automation does not change

De-duplication is automatic. You do not need to track which documents have already been processed — DASH does that per channel.
The first poll respects your start point. You choose whether the very first poll reaches back to older documents or only picks up new ones from now on. See Configure polling.
You can test before going live. A dry-run poll produces sample certificates you can review without affecting your real certificate list. See Run a dry-run.