Workflow builder demo

Build a PDF workflow from document to API calls.

Choose a public PDF, pick the outcome, and switch between a simple path and a complex production-style workflow.

Generated plan

A governed intake path extracts data, flags low confidence cells, routes review, and exports evidence.

Document

Form W-9

Mode

complex

API calls

Plan ready

4 steps

Artifacts

4 outputs

Recipe

TypeScript

1. Public PDF

2. Desired outcome

3. Workflow depth

Complex paths add review gates, evidence, and downstream integration.

4. API language

Node 20+, Bun, or any server runtime with fetch and FormData support.

app.docuzeus.local/extract-tables

Data workspace

Active tool

Extract Tables

Tables: 8
Fields: 42
Needs QA: 3

Adobe FY2025 Form 10-K

U.S. Securities and Exchange Commission EDGAR

Processing

Table region

Low confidence cell

CSV ready

Structured output

Detected rows, fields, and confidence flags are prepared for CSV or JSON export.

Human review

Low-confidence cells are grouped before downstream automation consumes them.

Pipeline ready

Extraction can run as a step after OCR or before AI summary.

Document set

Form W-9

6 pages

Adobe FY2025 Form 10-K

134 pages

Attention Is All You Need

15 pages

Extracted values

Tables8

Cells flagged3

SchemaMapped

Form W-9complex workflow

Extract structured data

Best for field extraction, metadata cleanup, secure intake, and redaction practice.

Also show the simple path

One upload becomes a searchable PDF plus exported fields.

Generated tool chain

Upload with workspace metadata.
Run OCR, extraction, and field normalization as a pipeline.
Review low-confidence values before export.
Attach final artifacts to the source record.

API call sequence

POST /v1/documents/uploadPOST /v1/pipelines/run-adhocGET /v1/jobs/{job_id}/statusGET /v1/convert/jobs/{job_id}/artifacts

Integration overview

Expected outputs

Normalized metadata
Reviewed data package
Pipeline run history
Audit-ready artifacts

Linked tools

Forms Ops Studio OCR Extract Tables Set Metadata Studio Pipeline Builder

Run feedback

Queued

Document accepted.

Running

4 API calls.

Review

5 tools linked.

Complete

4 artifacts ready.

TypeScript implementation preview

Secure W-9 intake packet / Form W-9 / complex

Integration overview

const apiBase = process.env.DOCUZEUS_API_URL!;
const key = process.env.DOCUZEUS_API_KEY!;

async function request(path: string, init: RequestInit = {}) {
  const res = await fetch(`${apiBase}${path}`, {
    ...init,
    headers: { Authorization: `Bearer ${key}`, ...init.headers },
  });
  if (!res.ok) throw new Error(await res.text());
  return res;
}

const form = new FormData();
form.set('file', new Blob([w9PdfBytes], { type: 'application/pdf' }), 'irs-w9.pdf');
const upload = await (await request('/v1/documents/upload', { method: 'POST', body: form })).json();

const pipeline = await (await request('/v1/pipelines/run-adhoc', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
    'Idempotency-Key': `w9-intake-${upload.document_id}`,
  },
  body: JSON.stringify({
    document_id: upload.document_id,
    steps: [
      { tool: 'ocr' },
      { tool: 'extract-fields', profile: 'taxpayer-intake' },
      { tool: 'redact', policy: 'tin-release-copy' },
      { tool: 'redaction-verify' },
      { tool: 'compress', quality: 'medium' },
    ],
  }),
})).json();

let status;
do {
  await new Promise((resolve) => setTimeout(resolve, 2000));
  status = await (await request(`/v1/jobs/${pipeline.job_id}/status`)).json();
} while (status.status === 'queued' || status.status === 'running');

if (status.status !== 'completed') throw new Error(`Pipeline failed: ${status.status}`);
const evidence = await request(`/v1/forensics/evidence/${pipeline.job_id}`);
await writeFile('w9-evidence.zip', Buffer.from(await evidence.arrayBuffer()));

Explore the workflow Open matching screenshot