Abstract: A method and apparatus for fraud detection during transactions using identity graphs are described. A method includes receiving, at a commerce platform system, a transaction from a user having initial transaction attributes and transaction data. The method also includes determining, by the commerce platform system, an identity associated with the user associated with additional transaction attributes not received with the transaction. Furthermore, the method includes accessing a feature set associated with the initial transaction attributes and the additional transaction attributes that includes machine learning (ML) model features for detecting transaction fraud.

Abstract: Presented herein are systems and methods of evaluating network operations associated with computing systems. A server may receive, from a computing system, an electronic request to execute a first network operation using a plurality of attributes provided by an end user device to the computing system. The first network operation may be initiated via the end user device. The server may retrieve (i) a digital fingerprint associated with an identity of the computing system and (ii) a plurality of network operation metrics associated with the computing system. The server may execute, using the digital fingerprint and the plurality of network operation metrics, a machine learning (ML) model to generate a likelihood of fraud caused by the computing system. The server may, in response to the likelihood of fraud satisfying a threshold, execute a second network operation using the plurality of attributes, instead of executing the first network operation.

Abstract: A method for de-duplicating data records may include generating a nodal data structure comprising a set of nodes, in response to receiving a first set of transaction data corresponding to aggregated data associated with a first set of transactions, associating a first subset of nodes with the first set of transaction data, responsive to receiving a second set of transaction data corresponding to aggregated data associated with a second set of transactions and at least a subset of the first set of transactions, associating a second subset of nodes to the first set of transaction data, combining the first subset and the second subset, executing an analytical protocol to identify data associated with the transaction, disassociating the transaction with the first subset, and associating the data associated with the transaction identified via the analytical protocol with the first subset and the second subset.

Abstract: A machine learning framework and method for using the same are described.

Abstract: A method and apparatus for machine learning model (MLM) based fraud detection of transactions are described. The method may include determining a first block rate of a first MLM that scores transactions, wherein the fraud detection system blocks a transaction when a score generated for the transaction by the first MLM does not satisfy a first threshold value. The method may also include accessing a second MLM to be used by the fraud detection system for blocking future transactions. Furthermore, the method may include determining a second threshold value using a set of prior transactions input into the second MLM that results in a second block rate of the second MLM within a predetermined margin of the first threshold. The method may also include deploying the second MLM using the second threshold value for performing fraud detection of at least one new transaction.

Abstract: A method and system for performing a movement of a service within a distributed system. In one embodiment, the method comprises determining one or more locations where the service is to be moved. The method further comprises for each of the one or more locations, computing a metric including a cross-cell cost (C-cost) or a data locality-cost (DL-cost) of the software service if a movement of the software service to a location was to occur, prior to performing the movement of the software service; and modelling a latency and availability of the system due to the movement of the software service to the location based on the metric. The method further comprises performing the movement of the service based on the modelled latency and availability of the system.

Abstract: A method comprising: creating, on a server, an object corresponding to the transaction; tracking state of the object as the object transitions between a plurality of states, the object being in one state while waiting for a transfer for the transaction; and automatically associating with the object by matching a reference code received with the transfer.

Abstract: A method and apparatus for generating and using secure sharded onboarding user interfaces are described. The method may include determining, based on account type of a merchant account being onboarded for a merchant by a first user, that information for a user type associated with a second user associated with the merchant is to be collected to satisfy minimum merchant account compliance requirements. The method may also include allocating a person object in an accounts data store maintained by the commerce platform system, the person object associated in the accounts data store with the merchant account. Furthermore, the method can include generating a secure link and transmitting the secure link to the second user, the secure link resolving at the commerce platform system and comprising a unique identifier that is associated with the person object.

Abstract: A method and apparatus are described for modeling a software system of a distributed services system. The method includes receiving a stream of ledger transaction messages generated by a software system, which are then stored in a data store. A ledger system accesses a data model representing a state machine of the proper functioning of the software system. A ledger clearance operation is performed using the state machine and the data stored within the stream of ledger transaction messages in the data store to generate balances within states of the state machine. The ledger system detects an operational characteristic, such as proper function or improper function, of the software system based on a balance of a state in the state machine. Such detection is performed without knowledge of the underlying implementation of the software system.

Abstract: A server computer system is configured to forward secure data. The system stores secure data in a database. Each unit of secure data for which the system supports forwarding is associated with a respective token. The system receives from a first server, a request comprising at least: an indicator associated with a secure data, a destination indicator, a first API key, a second API key, and a body having a predefined data format. The system validates the request based on the first API key, references the database to obtain the secure data based on the indicator and the respective token, and populates the body with the secure data. Once populated, the system transmits a second request comprising the second API key and the populated body to a second server, based on the destination indicator.

Abstract: A method and apparatus for fraud detection during transactions using identity graphs are described. The method may include receiving a document image for detecting whether an identity document depicted within the document image is fraudulent. The method may also include extracting data associated with the document image to generate extracted data comprising image data extracted from the document image, image file data extracted from an image file for the document image, or a combination thereof. The method may also include processing, by a set of machine learning models, corresponding subsets of the decoded image data used as input to each machine learning model of the set of machine learning models, and further by a second machine learning model that generates a final score indicative of whether the document image depicts a fraudulent identity document, at least one or more initial scores.

Abstract: A method and apparatus for leveraging a distributed services system for surfacing insights for a user are described. The method may include accessing, by a computer processing system, service system data generated for a user by services of a distributed service system, the service system data including one or more values associated with one or more corresponding common entities in a data store of disambiguated service system data. The method may also include inputting the one or more values associated with the one or more corresponding common entities generated for the user into a trained machine learning model (MLM), the machine learning model trained to detect an effect of the one or more values associated with the one or more corresponding common entities on a set of metrics of interest. Furthermore, the method may include detecting, by the MLM, when there is an anomaly in a metric of interest in the set of metrics of interest.

Abstract: Aspects of the subject technology include determining an association between a first and second user account that comprises a first condition corresponding to a first action to be performed in association with the first user account and a second condition corresponding to a second action to be performed in association with the second user account. Aspects also include generating a first entry based on the association, receiving a first indication indicating completion the first and/or second actions, generating a second entry based on the first indication, wherein the second entry corresponds to the first and/or second conditions, determining whether the first and second condition of the first entry are satisfied based in part on entries in the data structure that correspond to the first and/or second conditions, and transmitting a second indication that the first entry is complete, in response to determining that the first and second condition are satisfied.

Abstract: Aspects of the subject technology include obtaining a transfer request event associated with a transaction, the transfer request event indicating a request day of week and time of day, and, when the request time of day is past a pre-determined cutoff time, adjusting the request day of week to be a subsequent day. Aspects also include determining, using the transfer request event and based on a machine learning model trained on historical transaction data, a predicted transfer delay for the transaction, obtaining a set of relevant non-transfer days based on a comparison between the historical transaction data and a set of past non-transfer days, and, when one or more non-transfer days from the set of relevant non-transfer days occur within a time period from the request day of week and over the predicted transfer delay, adjusting the predicted transfer delay based on the one or more non-transfer days.

Abstract: Discussed herein are methods and systems for dynamically recalibrating machine learning model parameters. In one method, a server executes one or more prediction models to process network operations from various data feeds, in order to identify the likelihood of these operations being fraudulent or malicious. The server monitors performance data, such as the operation and execution metrics of network operations, and evaluates whether the performance values, like recall values, meet defined thresholds. If the performance data fails to meet these thresholds, the server employs a function-generation machine learning model to predict a threshold modification function. This modification function is then applied to adjust the relevant thresholds. Utilizing the modification function, the server dynamically revises one or more parameters of the prediction models to enhance their accuracy and efficacy.

Abstract: Some embodiments involve systems and methods for facilitating authentication for an entity system to use a partner system using pooled identifiers (e.g., without requiring the entity system to first register with the partner system). In some embodiments, based on (i) a request, from a first entity system, indicating a first partner system to process at least a part of an operation and (ii) an indication that the first entity system is not registered for use of the first partner system, a platform system may link a first authentication identifier of the first identifier set with the first entity system. The platform system may process the operation using the first authentication identifier linked by the platform system with the first entity system, and the platform system may use platform authentication data established with the first partner system to authenticate the platform system to the first partner system for performing the operation.

Abstract: A method and apparatus for fraud detection during transactions using identity graphs are described. A method includes receiving, at a commerce platform system, a transaction from a user having initial transaction attributes and transaction data. The method also includes determining, by the commerce platform system, an identity associated with the user associated with additional transaction attributes not received with the transaction. Furthermore, the method includes accessing a feature set associated with the initial transaction attributes and the additional transaction attributes that includes machine learning (ML) model features for detecting transaction fraud.

Abstract: A transaction is conducted between a merchant site and a customer's electronic device using a payment processor. The merchant site is associated with a client-side application and a server-side application. The client-side application executes on the customer's electronic device. The client-side application electronically sends payment information retrieved from the customer's electronic device to the payment processor. The client-side application does not send the payment information to the server-side application. The payment processor creates a token from the payment information sent by the client-side application. The token functions as a proxy for the payment information. The payment processor electronically sends the token to the client-side application. The client-side application electronically sends the token to the server-side application for use by the server-side application in conducting the transaction.

Abstract: A system may include a network infrastructure having a set of network component nodes, each network component node configured to communicate with at least one other network component node in accordance with a dependency protocol; and a server in communication with the network infrastructure and a fault injection server. The server can be configured to monitor outputs generated by the network infrastructure and attributes of data communication between the set of network component nodes; execute a computer model using the dependency protocol and the monitored attributes and outputs as input to predict a set of faults; in response to presenting the set of faults for display on a user interface, receive a selection of one or more of the set of faults; and instruct the fault injection server to execute a fault injection scenario simulating performance of the network infrastructure operating under the selected one or more faults.

Abstract: In an example embodiment, a random forest machine learning algorithm is used to create and/or identify rules to apply to an individual entity in a computer system that has a plurality of entities, each with a number of rules. More precisely, rule predicates are used as features of a random forest model built to predict a particular outcome (e.g., a transaction that is fraudulent). Hyperparameters of the random forest model are varied and iterated. A classifier is used to calculate feature importance for all features in the training data. Feature importance may be calculated using permutation feature importance. The N “most important” features are then found from this set. The N “most important” features are then used to find rules above a certain precision and recall rate. These rules may then be backtested and the best rules can be used to generate additional rules.