Adding identity verification to an application in-house can be slow, laborious, and complex to implement. Not having proper identity collection and verification controls in place can increase the chances of fraud and account takeover, especially in industries such as fintech, banking, and healthcare. With 91% of organizations planning to increase spending on identity verification in the next three years, it’s vital for IT leaders and practitioners to think through this initiative.
Descope’s connector with Amazon Rekognition enables developers to add AI-powered facial recognition and identity verification to their authentication and user journey flows. Organizations using the connector can collect user IDs during registration, employ facial recognition to verify identities to reduce fraud, and lower user verification costs.
Visit our documentation for implementation details.
About Amazon Rekognition
Amazon Rekognition offers pre-trained and customizable computer vision capabilities to help extract information from images and videos. The service helps developers easily add computer vision APIs to their apps without needing to build ML models or infrastructure from scratch.
In an authentication context, Amazon Rekognition’s facial comparison and analysis can be used during onboarding or sign-in flows to accurately verify the identity of opted-in users. This can guard against fraudulent account openings or transactions, strengthening existing authentication flows.
The Descope connector with Amazon Rekognition enables developers to:
Collect user ID documents and register them after checking Amazon Rekognition confidence scores to ensure validity.
Verify a user’s identity by comparing a selfie taken during the authentication process with the stored image / ID of the user and checking Amazon Rekognition confidence scores.
Create branching user journey paths in Descope authentication flows based on the ID collection / verification confidence scores provided by Amazon Rekognition.
Use case: ID collection during sign-up
Many organizations struggle with fraudulent or spam account creation. Depending on the industry and nature of the app, one way to ensure clean, accurate, and legitimate user accounts is by collecting user ID during the onboarding process. Embedding the Amazon Rekognition “Register” action within Descope Flows makes this initiative straightforward.
The Flow screenshot below shows a sign-up process with email OTP as the authentication method. The user is then asked to upload an identification document, the validity of which is checked using an Amazon Rekognition action. Based on the documentConfidence score provided by Amazon Rekognition, branching user paths are created to continue onboarding.
The screen below shows an example of the signup experience:
Use case: Identity verification during sign-in
Facial recognition during sign-in can result in a powerful risk-based MFA approach that’s very hard for attackers to exploit. The Flow screenshot below shows a sign-in process that asks users to verify their identity by sharing a selfie. The selfie is compared with the existing ID document at hand for the user. Based on the externalIdMatched and confidence values from Amazon Rekognition, branching user paths are created to complete login or show a warning message.
The screen below shows an example of the identity verification process:
This connector combines the advanced computer vision capabilities of Amazon Rekognition with the drag-and-drop authentication and user journeys of Descope to help developers easily add identity verification and fraud prevention controls to their login flow.
Interested to learn more about the connector? Check out our docs. If you haven’t yet started your Descope journey, sign up for a Free Forever account and set up your own Amazon Rekognition connector. To explore other connectors, visit our integrations page.