Tutorial: Creating a Workflow

This tutorial walks you through creating a verification workflow from scratch using the visual graph builder. By the end, you’ll have a workflow that verifies a user’s identity with document scanning, facial recognition, age verification, and duplicate detection.

Prerequisites

• A Verifa account (sign up)
• Access to the dashboard in Sandbox mode

Step 1: Navigate to Workflows

From the dashboard, expand the Workflows section in the left sidebar under Configure, then click Workflows.

Step 2: Create a new workflow

You’ll see a list of your existing workflows. Your organization starts with a Default Workflow that handles basic ID scan, face match, and age verification.

Click + New Workflow in the top right corner to create a new one.

Step 3: Choose a starting point

A modal appears with template options. You can start from a pre-built template filtered by industry (Marketplace, Fintech, Age-Gated, Healthcare, Crypto) or start from scratch.

For this tutorial, select Blank Workflow to build everything step by step.

Step 4: Configure capture requirements

Before building the graph, Verifa asks what data the workflow should collect from applicants during the capture flow.

Enable selfie capture

Toggle on Selfie photo to require a live selfie for facial comparison. This is needed for the Facial Recognition check later.

Enable user information

Toggle on User information to collect personal details from applicants during the capture flow.

Select user info fields

After enabling User information, the field configuration expands. For this tutorial, enable the following fields and set them to required:

• First Name — collect, required
• Last Name — collect, required
• Email — collect, required
• Phone — collect, required

You can leave Date of Birth, SSN, and Address disabled for now — the document scan will extract date of birth automatically from the ID.

Click Continue to Builder when you’re done.

Step 5: Understand the graph builder

A quick tutorial modal explains how the workflow graph works. Every workflow starts with a Capture node at the top. From there, you build a chain of verification checks, branches, and outcomes. Data flows top-to-bottom through each node.

Think of it as a flowchart — each path represents a route an applicant can take through your verification process.

Click X to dismiss the tutorial (or click through all the tips — they’re helpful for first-time users).

Step 6: Add Document Scan

Click the + button below the Capture node to add your first verification step. The Add Step menu appears with all available checks organized by category:

• Extraction — Document Scan
• Verification — Facial Recognition, NFC Chip Authentication
• Evaluation — Address Validation, Age Verification, AML Screening, List Check, and more

Select Document Scan. This check extracts the applicant’s name, date of birth, address, and document number from their ID.

Step 7: Add Facial Recognition

Now add Facial Recognition after Document Scan. Click the + button on the pass output of the Document Scan node, then select Facial Recognition from the Add Step menu.

This check compares the applicant’s selfie against the photo on their ID document.

Step 8: Add Age Verification

Repeat the same process — click the + on the pass output of Facial Recognition and select Age Verification.

When you click on the Age Verification node, the configuration panel opens on the right side. Here you can set the minimum_age threshold. The default is 18, but you can change it to any value (e.g., 21 for alcohol delivery, 25 for car rentals).

Step 9: Add Duplicate Detection

Click the + on the pass output of Age Verification and select Duplicate Detection. This check searches for the same person across all previous sessions using hashed device fingerprints, email, phone, document number, and face embeddings.

Press Spacebar to auto-layout the graph after adding the node.

Step 10: Add Auto-Approve

Finally, click the + on the pass output of Duplicate Detection and select Auto-Approve. This is the decisioning step — if all prior checks passed, the session is automatically approved. If any check failed, the session routes to manual review or rejection.

Your complete workflow now looks like this:

Capture → Document Scan → Facial Recognition → Age Verification
  → Duplicate Detection → Auto-Approve

Each step also has a fail output that routes to Manual Review by default, so failed checks don’t silently disappear.

Step 11: Review capture requirements

You can review and edit the capture requirements at any time by clicking the Capture Requirements panel in the top left of the graph builder. This shows a summary of what documents, biometrics, and user info the workflow will collect.

There are multiple ways to open and edit these settings — you can click the panel header or use the Capture Requirements Settings button in the top toolbar.

Step 12: Name and save the workflow

Give your workflow a descriptive name by clicking the name field at the top of the graph builder (e.g., “Facial, Age, Duplicate”). Add an optional description.

Then click Save Workflow in the top right corner. This saves the workflow as a draft — it won’t process any sessions until you publish it.

Step 13: Review the workflow detail

After saving, you’re taken to the workflow detail page. Here you can see:

• Details — Status, publish status (Draft), version, and last updated timestamp
• Data Requirements — What the capture flow will collect (documents, selfie, user info)
• Auto-injected checks — Watchlist Screening, Risk Assessment, and Identity Cross-Reference
• Workflow steps — All 5 steps in order with their fail/review routing

Click Publish when you’re ready to make the workflow live.

Step 14: Publish the workflow

A confirmation modal appears showing the rollout percentage. This controls how much traffic the new version receives:

• 100% — All new sessions using this workflow will run on the new version immediately
• 50% — Half of sessions use the new version, half continue on the previous version
• 10% — Canary deployment — test with a small percentage before rolling out fully

Rollout percentage lets you safely deploy workflow changes in production. For a brand new workflow, leave it at 100% and click Publish v2.

Step 15: Workflow is live

Your workflow is now Published and ready to process sessions. The detail page shows:

• Publish Status: Published
• Version: v2
• Rollout: 100%

You can click Set as Default in the top toolbar to make this the default workflow for all new sessions, or specify it per-session by passing the workflow_id when creating a session via the API.

Using your workflow via the API

Now that your workflow is published, create sessions that use it:

$
curl -X POST https://api.withverifa.com/api/v1/sessions \
>
  -H "X-API-Key: vk_sandbox_your_key_here" \
>
  -H "Content-Type: application/json" \
>
  -d '{
>
    "external_ref": "user_abc123",
>
    "workflow_id": "wf_your_workflow_id"
>
  }'

Or set it as the default workflow and all new sessions will use it automatically.

What’s next

• Quickstart — Test your workflow with a sandbox session
• Workflows — Deep dive into graph structure and conditional routing
• Verifications & Checks — All available check types
• Cases — How to review sessions routed to manual review
• Fraud Signals — The 150+ signals collected during every session