👷Workflows

Validation workflows are a collection of tasks or steps organized in a structured sequence to effectively verify a person's identity. They can be customized upon request, however Tagshelf aims to provide the most common use cases as default options in our platform.

Default Workflows

The most recent version of our API contains two default endpoints available for our customers.

Signup Workflow

Use Case

This workflows is useful when you need to establish the identity of users who are new to your system. The workflow collects the user's ID information and performs a face match with their identity document. Real-life applications of this workflow are:

• InsureTech: A customer just became a member of your health insurance service and you need to collect their biometric face data and compare to a valid national identity document.

• Membership Validation: A customer just became a member in your gym and you need to collect an image of his or her face for security purposes.

Use this workflow when onboarding new users into your system, especially when you don't have prior knowledge of the users.

Required Fields

Field	Type	Description
face	File	This field expects binary data that represents the user's face image.
document	File	This field expects binary data that represents the user's document, which could be an identification card or passport.
documentSide	String	This field specifies whether the document provided is the "FRONT" or "BACK" side of the user's identification document. It ensures that the correct side is used for verification.
documentType	String	This field specifies the type of identification document being used. It can be "CEDULA" or "PASSPORT," indicating the document's category.
metadata	JSON String	Metadata is an object that may include additional information or context about the validation process. It can be used to provide extra details or annotations related to the validation request.
threshold	Number	The threshold is an optional numeric value used to determine the similarity or dissimilarity between the face and document images. It helps set a threshold for the comparison result. The threshold value can be customized according to the desired level of validation accuracy.

Validation Workflow

Use Case

This workflow is ideal for confirming the identity of a user who previously signed up. In this case, you only need to provide a face, and 2FACE compares this face with all previous faces associated with the user on our platform. It offers a quick and reliable way to validate your users and ensure the person using your app is who they claim to be. Some real-life examples of this workflow are:

• Banking and FInancial operations: many banks and financial institutions use facial recognition technology to authenticate users during online transactions. The validation workflow allows for this authentication to occur.

• Transportation: Transport companies increasingly use facial recognition systems for identity verification during check-in, boarding, and security checks.

Required Fields

Field	Type	Description
face	File	This field expects binary data that represents the user's face image.
documentId	String	This parameter expects a string value representing the Document ID that you want to validate the user's face against.
metadata	JSON String	Metadata is an object that may include additional information or context about the validation process. It can be used to provide extra details or annotations related to the validation request.
threshold	Number	The threshold is an optional numeric value used to determine the similarity or dissimilarity between the face and document images. It helps set a threshold for the comparison result. The threshold value can be customized according to the desired level of validation accuracy.

For a detailed description of this endpoint and definition of the job status, please visit Jobs

Custom Workflows

Custom worklflows are available upon request and design with Tagshelf's development team. If you or your team are interested in developing a custom 2FACE workflow, please send an email to [email protected]