LT Pipelines

LT (Language Technology) pipelines automate text and speech processing tasks on your files. Unlike Verification Pipelines, which produce an authenticity verdict, LT pipelines extract or transform content — returning transcripts, translations, detected languages, named entities, or extracted text.

Each pipeline runs on one file at a time. Results are stored as structured JSON on the run record and can be retrieved via the API or viewed inline in the platform.

Available Pipelines

How to Trigger a Pipeline

You can run an LT pipeline from the platform UI or via the API.

From the Platform

Open a project or dataset version, navigate to the LT Pipelines tab, select a pipeline, pick the files to process, fill in any required context fields, and click Run.

Via the API

On a project:

POST /v2/lt-pipelines/project/{projectId}/run

{
  "pipelineSlug": "lt-transcription",
  "fileIds": ["file_abc123"],
  "contextData": {}
}

On a dataset version:

POST /v2/lt-pipelines/dataset/{datasetId}/versions/{versionId}/run

{
  "pipelineSlug": "lt-translation",
  "contextData": { "target_language": "English" }
}

The dataset endpoint runs the pipeline on every file in the version automatically. Both endpoints return a runIds array — one ID per file.

Checking Results

Poll or fetch a single run:

GET /v2/lt-pipelines/runs/{runId}

List all runs for a project:

GET /v2/lt-pipelines/project/{projectId}/runs

You can filter by status (pending, running, completed, failed) as a query parameter.

Pipeline Pages