> ## Documentation Index
> Fetch the complete documentation index at: https://vaquill.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart: your first workflow

> Go from a folder of contracts to a finished review memo in about ten minutes

The fastest way to understand workflows is to run one. This walks through [NDA Bulk Review](/docs/workflows/built-in/nda-bulk-review), which needs nothing but a few NDAs and about ten minutes.

<Frame caption="The Workflows gallery, where you pick a workflow to run">
  <img src="https://mintcdn.com/vaquill/2UMb7US0EYXDhATr/images/workflows-gallery.webp?fit=max&auto=format&n=2UMb7US0EYXDhATr&q=85&s=99b8670d4e330b12384ea7bd400b9c8f" alt="Screenshot of the Workflows gallery showing the built-in catalogue you launch a run from" style={{ borderRadius: '0.5rem' }} width="1712" height="1070" data-path="images/workflows-gallery.webp" />
</Frame>

## Before you start

<Steps>
  <Step title="An active account">
    A [trial or paid plan](/docs/guides/plans-and-limits). Workflows are included in every plan and are not billed per run.
  </Step>

  <Step title="A few documents">
    Two to five NDAs in PDF or DOCX. **They must be text-searchable.** A scanned image runs through OCR on upload, which works, but give it time to finish before launching.
  </Step>

  <Step title="Optionally, a matter">
    You can run a workflow without one. Scoping to a matter files the output back into that matter automatically, which is worth doing on real work. See [Matters and workspaces](/docs/guides/matters-workspaces).
  </Step>
</Steps>

## Run it

<Steps>
  <Step title="Open the catalogue">
    Go to **Workflows** in the app and pick **NDA Bulk Review** from the In-House group.
  </Step>

  <Step title="Upload into the slot">
    The launcher shows a slot labelled **NDAs to review**. Drag your files in. Leave the optional playbook slot empty for now: with no playbook, the review scores against market benchmarks instead.
  </Step>

  <Step title="Answer one required question">
    **Which side are you on?** is the only required field. It inverts most of the scoring rubric, so pick honestly: Receiving Party if you are the one being bound.

    Everything else is optional and falls through to sensible defaults. Ignore it on the first run.
  </Step>

  <Step title="Launch, then walk away">
    Click run. The pipeline builds an extraction matrix, answers a cell per document per clause question, scores each NDA, then synthesizes across the set. Expect 3 to 12 minutes.

    You do not need to keep the tab open. Progress streams into the run view, and you can cancel at any point.
  </Step>

  <Step title="Collect two documents">
    On completion you get an **Executive Summary** with a traffic-light grid across every NDA, and a **Per-Document Detail** with benchmark results and suggested redline language. Both are real DOCX files.
  </Step>
</Steps>

<Tip>
  Uploaded and launched straight away, and retrieval came back thin? Documents have to finish ingesting before a workflow can read them. Give a fresh upload a minute.
</Tip>

## What just happened

You ran a fixed pipeline someone else designed. Every [built-in workflow](/docs/workflows/built-in) works the same way: slots for documents, a short form that changes what counts as risky, and a declared set of output files.

That is the whole model. The 24 other built-ins differ in what they extract and what they produce, not in how you drive them.

## Where to go next

<CardGroup cols={2}>
  <Card title="Pick a workflow for your actual job" icon="grid-2" href="/docs/workflows/built-in">
    The full catalogue, grouped by the work: contracts, compliance, research, discovery, drafting.
  </Card>

  <Card title="Learn the launcher properly" icon="play" href="/docs/workflows/running-a-workflow">
    Auto-fill from a document or brief, matter scope, progress, and cancelling.
  </Card>

  <Card title="Build your own" icon="wrench" href="/docs/workflows/builder">
    When no built-in matches, chain 25 step types into your own pipeline, or describe it in plain English.
  </Card>

  <Card title="Start from a template" icon="copy" href="/docs/workflows/builder/templates">
    Eight ready-made custom workflows. Document Triage is three steps and a good one to fork first.
  </Card>
</CardGroup>

## If something goes wrong

| Symptom                             | Cause                                                                                                      |
| ----------------------------------- | ---------------------------------------------------------------------------------------------------------- |
| The workflow will not launch        | A required slot is empty or below its minimum document count. The launcher names the slot.                 |
| Retrieval finds nothing             | The documents were still ingesting. Wait, then re-run.                                                     |
| Everything returns a payment error  | The trial lapsed. See [Plans, trial, and limits](/docs/guides/plans-and-limits).                                |
| The output is thinner than expected | A scanned document that OCR'd poorly, or the wrong side selected on the "which side are you on?" question. |

For builder-specific errors, see [Troubleshooting](/docs/workflows/builder/troubleshooting).