How long uploaded files and extraction results are kept. Per-artifact, set at submission time.

Every artifact carries a retention_policy that controls how long the uploaded file and its extraction results live in storage. The default is 30 days. Set it at submission time per-artifact, or globally per pipeline.

Available policies

The artifact record (metadata, status, extracted JSON result) is retained longer than the source file in most policies; typically until you delete the artifact explicitly. Only the raw upload is purged on the schedule.

Setting a policy

Pass retention_policy on submit:

{
  "filename":         "passport-scan.png",
  "artifact_type":    "image/png",
  "artifact_size":    921384,
  "retention_policy": "immediate",
  "template_id":      "tpl_id_card"
}

Or as a form field on POST /artifacts/upload:

curl -u "$DD_KEY:$DD_SECRET" \
  -X POST https://api.datadistillers.com/api/v1/artifacts/upload \
  -F 'file=@./passport.png' \
  -F 'retention_policy=immediate'

The chosen policy is stored on the artifact and visible in subsequent GET /artifacts/{id} responses. It cannot be changed after submission; delete and re-submit if you need a different policy.

Picking a policy

A useful starting point keyed by document sensitivity:

For regulated workloads (HIPAA, GDPR's "right to be forgotten"), prefer the shortest policy that still allows your processing window, and pair it with explicit DELETE /artifacts/{id} calls when a user requests removal.

Retention vs explicit delete

Retention is the safety net. Explicit DELETE /artifacts/{id} is the authoritative removal; it deletes the source file, the result, and the artifact record itself, all at once.

curl -u "$DD_KEY:$DD_SECRET" \
  -X DELETE https://api.datadistillers.com/api/v1/artifacts/art_b4d8c0f1

Returns 204 No Content on success. There's no soft delete; the record is gone the moment the call returns. Orchestrate this from your application when a user-initiated deletion event fires (account closure, GDPR request, etc.); don't rely on retention alone.

What happens after the policy expires

Once the source file is purged:

• GET /artifacts/{id}/download?file_type=raw returns 404.
• GET /artifacts/{id}/download?file_type=json still works as long as the artifact record exists.
• POST /artifacts/{id}/rerun returns 400; there's no source to re-process.
• Job status APIs continue to return historical data.

If you anticipate needing to re-process, choose a longer retention policy upfront. There's no recovery once the file is gone.