KYCFeature Guides

IDV & Biometrics

Learn how to implement IDV and Biometrics in your KYC processes.

Introduction

FrankieOne’s Identity Verification (IDV) and Biometrics services are designed to accurately identify and onboard genuine individuals while effectively preventing fraud from falsified or stolen identification documents.

By integrating advanced document analysis, Optical Character Recognition (OCR), and biometric checks like facial comparison and liveness detection, you can build a robust, secure, and compliant onboarding workflow.

This guide will walk you through the end-to-end process of implementing our IDV and Biometrics solution, from creating an entity to initiating the verification workflow.

Key Features

  • Global Document Verification: Validate a wide range of government-issued identity documents from around the world.
  • Advanced Biometric Authentication: Secure your platform with facial recognition and liveness detection to prevent spoofing and ensure the user is physically present.
  • Automated Data Extraction (OCR): Reduce manual entry and errors by automatically extracting information from identity documents.
  • Configurable Workflows: Tailor the verification steps and logic to match your specific risk appetite and compliance requirements.
  • Streamlined Integration: Use our Hosted URL for the quickest integration or leverage our SDKs for a more customized user experience.

The End-to-End IDV Flow

The verification process involves a sequence of API calls to create an entity, initiate a capture session, process the captured data, and finally execute a workflow to get a result.

1

Step 1: Create an Individual Entity

Before you can perform an IDV check, you need an entityId for the user. If you don’t already have one, create an individual entity by making a server-side call to the API.

$POST /v2/individuals

This request creates the user’s profile in the FrankieOne system and returns the essential entityId that will be used in all subsequent steps.

2

Step 2: Initiate the Capture Session

Once you have an entityId, you can start the data capture process where the user will submit their document and a selfie. We offer two primary integration paths.

The simplest way to integrate is by using our Hosted URL. This provides a secure, FrankieOne-hosted webpage that guides the user through the entire capture process.

Make a server-side call to generate a unique, single-use URL for the user’s session.

$POST /v2/individuals/hostedUrl

You can then redirect the user to this URL. You can even configure it to send the link directly to the user via SMS by setting sendSMS to true in your request.

Option B: Direct SDK Integration (Advanced)

For a more deeply integrated, native experience within your own application, you can use an IDV provider’s SDK. To initialize it, you first need to retrieve a session token from our API.

$POST /v2/individuals/{entityId}/actions/idv/token

This token authorizes the provider’s SDK to conduct the capture session on behalf of the specified entityId.

3

Step 3: Trigger Data Processing

After the user successfully completes the document and selfie capture in the Hosted Flow or provider SDK, your backend must instruct FrankieOne to retrieve and process the captured data.

$POST /v2/individuals/{entityId}/actions/idv/process

This call signals that the capture is complete and triggers FrankieOne to fetch the results from the IDV vendor, including the OCR data and biometric information.

4

Step 4: Execute Workflow and Get Results

Upon successful data processing, the IDV service will automatically execute the workflow configured for the entity’s service profile (e.g., KYC). This workflow runs the necessary verification steps and generates the final results.

To learn how to analyze the detailed outcomes of these checks, please see our dedicated guide: Understanding IDV & Biometrics Results.

Understanding the Verification Checks

The IDV workflow orchestrates several powerful checks to ensure the integrity of the verification.

  • Document Authenticity: The system analyzes the document for signs of tampering, fraud, and inconsistencies, verifying security features and comparing it against known templates. Checks include visual_authenticity and image_integrity.
  • Biometric Verification: This involves two key checks:
    • Facial Liveness: Confirms the user is a real, live person and not a spoof attempt (e.g., a photo of a photo or a deepfake).
    • Facial Comparison: Matches the user’s live selfie against the photo on the identity document with a quantifiable confidence score.
  • OCR Data Comparison: After extracting data from the document via OCR, the system compares it against the information stored in the user’s entity profile. This crucial step flags discrepancies between what the user provided and what is on their official document.

FrankieOne uses fuzzy matching logic (the Levenshtein distance algorithm) to handle minor OCR errors or typos in names and numbers. This reduces false negatives from insignificant discrepancies. For example, by default, a name is allowed a Levenshtein distance of 2 before it is flagged as SUSPECTED.