USuseboom
Initiatives

Re-extract conversations

Changed on

Re-run extraction over an initiative's existing conversations — the fix for participants whose replies were never read (a schema authored after the conversations happened, or an initiative that never had extraction run at all). Start with dryRun to see how many conversations are eligible and how many actually carry enough transcript to extract from. Only conversations with at least 3 messages including one reply can produce anything.

post/api/v1/initiatives/{id}/reextract

Request

  • Base URL
    https://www.useboom.ai
  • Auth
    HTTP bearer

Path parameters

idstring required

The initiative id.

Request body

dryRunboolean

Count what would run without queueing anything. Do this first — the counts tell you whether re-extraction is worth the spend.

forceboolean

Re-extract conversations already extracted at the current schema version. Off by default, so a plain call only fills gaps.

fromDatestring date-time

Only consider conversations created at or after this ISO 8601 instant.

environmentIdstring

Which environment to re-extract. Optional only while the organization has exactly one; with more than one it is required, because this rewrites values in exactly one environment and defaulting could rewrite production from a staging test.

Response

Success

initiativeIdstring required
eligibleinteger required

Conversations targeted for re-extraction.

queuedinteger required

Extractions actually enqueued — always 0 on a dry run.

willExtractinteger required

Of the eligible set, how many carry enough transcript to produce values.

insufficientinteger required

Eligible but below the analyzability bar (under 3 messages, or no reply on this initiative) — these will be marked INSUFFICIENT_DATA rather than extracted. A high number here usually means the replies live on a different initiative.

dryRunboolean required

Changes