IDPForgeDocs
Open app
DocsUpdated 23 Aug 2026

Turn documents into data you can trust

Configure a pipeline, wire the review loop your team can actually hold, and call it all from the API. Start with the five-minute quickstart.

QuickstartPipeline anatomyAPI referenceChangelog

Start where you are

Developers

Submit jobs, read results, subscribe to events. REST, signed webhooks, SDKs for Python and TypeScript.

Quickstart in 5 minutesAuth & scopesWebhook signing
Get started →

Builders & ops

Design the pipeline: sources, classification, schemas, validation rules and destinations.

Pipeline anatomyTaxonomies & routingDeliver to a destination
Build a pipeline →

Reviewers

Clear exceptions, work a queue, and understand what the system records while you correct.

Correct a documentKeyboard shortcutsWhen to park
Work a queue →

Evaluating IDPForge

End-to-end scenarios with real numbers — including what went wrong and how it was fixed.

AP invoice automationMailroom intakeGlossary
Read a scenario →

Most read

Quickstart
Post a PDF, read the JSON back.
Exceptions & human review
Triggers, queues, profiles, coverage.
Submit a job
POST /v1/jobs, every parameter.
Pipeline anatomy
What each of the six stages owns.
Webhooks
Event catalogue and signature verification.
Glossary
The vocabulary that wins.

Your first extraction

One call submits a document to a published pipeline. The response is a job id; results arrive asynchronously or by webhook.

Open the quickstart →
curl https://api.idpforge.ai/v1/jobs \
  -H "X-API-Key: $IDP_API_KEY" \
  -F "pipeline_id=std-invoice" \
  -F "file=@invoice.pdf"
job = client.jobs.create(
    pipeline_id="std-invoice",
    file=open("invoice.pdf", "rb"),
)
const job = await idp.jobs.create({
  pipelineId: "std-invoice",
  file: fs.createReadStream("invoice.pdf"),
});