Flexible Identity and Access Management Pipeline

Abstract

A system and a method are disclosed for context-based verification flows for digital identity verification. A context-based verification system provides flexible identification procedures for various enterprises that are adapted to the enterprises' services, the enterprises' customers, and these customers' needs. For example, the context-based verification system determines first and second verification flows associated with a first enterprise and a third verification flow associated with a second enterprise. These verification flows include context parameters and verification parameters. The context-based verification system determines context parameters of a request, or “request context parameters,” when a user requests to interact with the first enterprise and determines a verification flow associated with context parameters that substantially match these request context parameters. The context-based verification system may determine that the first verification flow's context parameters substantially match the request context parameters and use the first verification flow's verification parameters to verify the user's identity.

What is claimed is: 
     
         1 . A computer-implemented method performed by a verification system, the method comprising:
 determining a plurality of verification flows associated with a first enterprise, wherein each verification flow, of the plurality of verification flows, is associated with a set of context parameters and a set of verification parameters;   determining one or more request context parameters associated with a request, by a user, to interact with the first enterprise; and   based at least in part on determining that the one or more request context parameters substantially match a first set of context parameters, of the set of context parameters, associated with a first verification flow of the plurality of verification flows, verifying an identity of the user using a first set of verification parameters associated with the first verification flow.   
     
     
         2 . The computer-implemented method of  claim 1 , further comprising:
 causing output of a graphical user interface comprising at least one first listing comprising a plurality of context parameters and at least one second listing comprising a plurality of verification parameters;   receiving, via the graphical user interface, an indication of a selection of the first set of context parameters from the plurality of context parameters and an indication of a selection of the first set of verification parameters from the plurality of verification parameters; and   generating, based at least in part on associating the selected first set of context parameters with the selected first set of verification parameters, the first verification flow.   
     
     
         3 . The computer-implemented method of  claim 1 , further comprising:
 receiving, from a third-party service, an evaluation of at least one of the first enterprise, the request to interact with the first enterprise, the user, or a user device associated with the user, wherein the evaluation comprises a risk assessment score,   wherein determining that the one or more request context parameters substantially match the first set of context parameters associated with the first verification flow is further based at least in part on the evaluation, wherein the one or more request context parameters comprises the risk assessment score.   
     
     
         4 . The computer-implemented method of  claim 1 , further comprising:
 generating, based at least in part on a machine learning model, the first verification flow, wherein the machine learning model is configured to receive as input a plurality of enterprise attributes and output at least one recommended verification flow.   
     
     
         5 . The computer-implemented method of  claim 4 , further comprising:
 updating, based at least in part on interaction information characterizing an interaction of the user with the first enterprise according to the request, a training set; and   re-training, using the updated training set, the machine learning model.   
     
     
         6 . The computer-implemented method of  claim 4 , further comprising:
 re-training, based at least in part on a measure of user satisfaction, the machine learning model, wherein the re-training comprises adjusting, based at least in part on the measure of user satisfaction, a strength of an association between the plurality of enterprise attributes and the at least one recommended verification flow.   
     
     
         7 . The computer-implemented method of  claim 1 , further comprising:
 encoding, into a first feature vector, a plurality of enterprise attributes characterizing the first enterprise;   determining, based on the first feature vector, a similar enterprise characterized by at least one of the plurality of enterprise attributes characterizing the first enterprise; and   generating, based at least in part on a verification flow used by the similar enterprise, the first verification flow for the first enterprise.   
     
     
         8 . The computer-implemented method of  claim 7 , wherein determining, based on the first feature vector, the similar enterprise comprises:
 encoding, into a second feature vector, a plurality of enterprise attributes characterizing the similar enterprise; and   calculating a cosine similarity between the first feature vector and the second feature vector.   
     
     
         9 . The computer-implemented method of  claim 1 , wherein determining the one or more request context parameters associated with the request comprises:
 determining, from the request, at least one of a target resource, an organization of the user, a device serial number, a device model number, a universally unique identifier (UUID), geographic coordinates, an IP address, or a MAC address.   
     
     
         10 . The computer-implemented method of  claim 1 , wherein determining whether the one or more request context parameters substantially match the first set of context parameters associated with the first verification flow comprises:
 determining a second set of context parameters associated with a second verification flow of the plurality of verification flows associated with the first enterprise;   determining, based at least in part on a first set of weights associated with the first set of context parameters, a first weighted score associated with the first verification flow;   determining, based at least in part on a second set of weights associated with the second set of context parameters, a second weighted score associated with the second verification flow; and   determining that the first weighted score associated with the first verification flow is greater than the second weighted score associated with the second verification flow.   
     
     
         11 . The computer-implemented method of  claim 1 , wherein verifying the identity of the user using the first set of verification parameters comprises:
 providing, to a secured account of the user, a unique verification identifier, wherein the unique verification identifier is at least one of a uniform resource locator (URL) or matrix barcode; and   based at least in part on receiving information indicating that the user has interacted with the unique verification identifier, fulfilling the request.   
     
     
         12 . The computer-implemented method of  claim 1 , wherein verifying the identity of the user using the first set of verification parameters comprises:
 receiving a first identification verifier;   based at least in part on verifying validity of the first identification verifier, providing the user access to a first resource, wherein the first resource fulfills a first portion of the request;   receiving a second identification verifier; and   based at least in part on verifying validity of the second identification verifier, providing the user access to a second resource, wherein the second resource fulfills a second portion of the request.   
     
     
         13 . A verification system comprising:
 a processor; and   memory storing instructions that, when executed by the processor, cause the verification system to:
 determine a plurality of verification flows associated with a first enterprise, wherein each verification flow, of the plurality of verification flows, is associated with a set of context parameters and a set of verification parameters; 
 determine one or more request context parameters associated with a request, by a user, to interact with the first enterprise; and 
 based at least in part on determining that the one or more request context parameters substantially match a first set of context parameters, of the set of context parameters, associated with a first verification flow of the plurality of verification flows, verify an identity of the user using a first set of verification parameters associated with the first verification flow. 
   
     
     
         14 . The verification system of  claim 13 , wherein the instructions, when executed by the processor, further cause the verification system to:
 cause output of a graphical user interface comprising at least one first listing comprising a plurality of context parameters and at least one second listing comprising a plurality of verification parameters;   receive, via the graphical user interface, an indication of a selection of the first set of context parameters from the plurality of context parameters and an indication of a selection of the first set of verification parameters from the plurality of verification parameters; and   generate, based at least in part on associating the selected first set of context parameters with the selected first set of verification parameters, the first verification flow.   
     
     
         15 . The verification system of  claim 13 , wherein the instructions, when executed by the processor, further cause the verification system to:
 receive, from a third-party service, an evaluation of at least one of the first enterprise, the request to interact with the first enterprise, the user, or a user device associated with the user, wherein the evaluation comprises a risk assessment score,   wherein determining that the one or more request context parameters substantially match the first set of context parameters associated with the first verification flow is further based at least in part on the evaluation, wherein the one or more request context parameters comprises the risk assessment score.   
     
     
         16 . The verification system of  claim 13 , wherein the instructions, when executed by the processor, further cause the verification system to:
 generate, based at least in part on a machine learning model, the first verification flow;   update, based on interaction information characterizing an interaction of the user with the first enterprise according to the request, a training set; and   re-train, using the updated training set, the machine learning model.   
     
     
         17 . The verification system of  claim 13 , wherein the instructions, when executed by the processor, further cause the verification system to:
 generate, based at least in part on a machine learning model, the first verification flow, wherein the machine learning model is configured to receive as input a plurality of enterprise attributes and output at least one recommended verification flow; and   re-train the machine learning model by adjusting, based at least in part on a measure of user satisfaction, a strength of an association between the plurality of enterprise attributes and the at least one recommended verification flow.   
     
     
         18 . The verification system of  claim 13 , wherein the instructions, when executed by the processor, further cause the verification system to:
 encode, into a first feature vector, a plurality of enterprise attributes characterizing the first enterprise;   determine, based on the first feature vector, a similar enterprise characterized by at least one of the plurality of enterprise attributes characterizing the first enterprise; and   generate, based at least in part on a verification flow used by the similar enterprise, the first verification flow for the first enterprise.   
     
     
         19 . The verification system of  claim 13 , wherein to determine whether the one or more request context parameters substantially match the first set of context parameters associated with the first verification flow the instructions, when executed by the processor, cause the verification system to:
 determine a second set of context parameters associated with a second verification flow of the plurality of verification flows associated with the first enterprise;   determine, based at least in part on a first set of weights associated with the first set of context parameters, a first weighted score associated with the first verification flow;   determine, based at least in part on a second set of weights associated with the second set of context parameters, a second weighted score associated with the second verification flow; and   determine that the first weighted score associated with the first verification flow is greater than the second weighted score associated with the second verification flow.   
     
     
         20 . A non-transitory, computer-readable storage medium comprising instructions that, when executed by a processor of a verification system, perform actions comprising:
 determining a plurality of verification flows associated with a first enterprise, wherein each verification flow, of the plurality of verification flows, is associated with a set of context parameters and a set of verification parameters;   determining one or more request context parameters associated with a request, by a user, to interact with the first enterprise; and   based at least in part on determining that the one or more request context parameters substantially match a first set of context parameters, of the set of context parameters, associated with a first verification flow of the plurality of verification flows, verifying an identity of the user using a first set of verification parameters associated with the first verification flow.