Abstract: There are provided systems and methods for a compute platform for machine leaning model roll-out. A service provider, such as an electronic transaction processor for digital transactions, may provide intelligent decision-making through decision services that execute machine learning models. When deploying or updating machine learning models in these engines and decision services, a model package may include multiple models, each of which may have an execution graph required for model execution. When models are tested from proper execution, the models may have non-performant compute items, such as model variables, that lead to improper execution and/or decision-making. A model deployer may determine and flag these compute items as non-performant and may cause these compute items to be skipped or excluded from execution. Further, the model deployer may utilize a pre-production computing environment to generate the execution graphs for the models prior to deployment or upgrading.

Abstract: An edge computing system is deployed at a physical location and receives an input from one or more image/video sensing mechanisms. The edge computing system executes artificial intelligence image/video processing modules on the received image/video streams and generates metrics by performing spatial analysis on the images/video stream. The metrics are provided to a multi-tenant service computing system where additional artificial intelligence (AI) modules are executed on the metrics to execute perception analytics. Client applications can then be run on the output of the AI modules in the multi-tenant service computing system.

Abstract: Techniques are disclosed for time constrained electronic request evaluation. A server system receives, from a computing device, a request submitted via an account, including a first set of characteristics associated with the request. The system executes a first machine-learning model to determine a first risk score for the request by inputting the first set of characteristics into the first model. The system generates an initial authentication decision for the request based on the first score and sends the decision to the device. The system executes a second, different machine-learning model to determine a second risk score for the request, by inputting the first set of characteristics and a second, different set of characteristics associated with the account into the second model. Based on the second score, the system determines a final authentication decision. The disclosed techniques may advantageously improve computer security and operations via identification of malicious electronic requests.

Abstract: Systems, methods, and computer program products are directed to machine learning techniques that use a separate embedding layer. This can allow for continuous monitoring of a processing system based on events that are continuously generated. Various events may have corresponding feature data associated with at least one action relating to a processing system. Embedding vectors that correspond to the features are retrieved from an embedding layer that is hosted on a separate physical device or a separate computer system from a computer that hosts the machine learning system. The embedding vectors are processed though the machine learning model, which may then make a determination (e.g. whether or not a particular user action should be allowed). Generic embedding vectors additionally enable the use of a single remote embedding layer for multiple different machine learning models, such as event driven data models.

Abstract: An edge computing system is deployed at a physical location and receives an input from one or more image/video sensing mechanisms. The edge computing system executes artificial intelligence image/video processing modules on the received image/video streams and generates metrics by performing spatial analysis on the images/video stream. The metrics are provided to a multi-tenant service computing system where additional artificial intelligence (AI) modules are executed on the metrics to execute perception analytics. Client applications can then be run on the output of the AI modules in the multi-tenant service computing system.

Abstract: This application discusses identifying data processing timeouts in live risk analysis systems. A service provider, such as an electronic transaction processor, may provide a production computing environment that includes a risk analysis system having one or more risk models, which may be machine-learning based. These risk models may be utilized in order to determine whether incoming data processing requests are fraudulent. To test these risk models using production data traffic, an audit computing environment made of a set of machines that do not service production computing environment requests, but that utilize databases and data connections as are used by the production systems. The audit computing environment may thus mirror the risk models and functionality of the production computing environment without the drawbacks of a typical fully separate testing environment.

Abstract: Described herein are various technologies pertaining to delivery of token-authenticated encrypted data. Content descriptor(s) (e.g., playlist(s)) can be modified to facilitate exchange of a token for a decryption key for browser(s) that do not provide logic to manage a flow of the token.

Abstract: A transaction management service and method are disclosed that can streamline transaction tracking and budget management. Transactions may be pre-approved based on inheritance and/or context-based authorization. The system can integrate a payment method, transaction tracking service, and a budget allocation service, among other operations.

Abstract: A sparse streaming system provides a first-class means for sparse metadata to be added to streaming media presentations and to be delivered using an integrated data channel that is cacheable using readily available HTTP-based Internet caching infrastructure for increased scalability. The sparse streaming system stores a reference to a sparse track within a continuous track. If a continuous fragment arrives at the client that refers to a sparse fragment that the client has not yet retrieved, then the client requests the sparse fragment. In addition, each sparse fragment may include a backwards reference to the sparse fragment created immediately prior. The references in the continuous fragments make the client aware of new sparse track fragments, and the backwards references in the sparse track fragments ensure that the client has not missed any intervening sparse track fragments.

Abstract: Described herein are various technologies pertaining to delivery of token-authenticated encrypted data. Content descriptor(s) (e.g., playlist(s)) can be modified to facilitate exchange of a token for a decryption key for browser(s) that do not provide logic to manage a flow of the token.

Abstract: Described herein are various technologies pertaining to delivery of token-authenticated encrypted data. Content descriptor(s) (e.g., playlist(s)) can be modified to facilitate exchange of a token for a decryption key for browser(s) that do not provide logic to manage a flow of the token.

Abstract: A coordinated adaptive streaming system provides a level of intelligence between adaptive streaming heuristics applied to multiple multi-bitrate streams playing on a client computer at the same time. The system receives a media event that includes multiple multi-bitrate streams and plays two or more of the streams. Each stream registers with the coordinated adaptive streaming system so that the system is aware of each of the streams. The system receives a priority indication from each stream that indicates the priority of the stream relative to other streams associated with the media event. The system uses the received priority indications to make adaptive streaming decisions. Thus, the coordinated adaptive streaming system provides playback that more closely aligns with a user's priorities and provides a higher quality experience to a user viewing multiple concurrent media streams.

Abstract: Described herein are various technologies pertaining to delivery of token-authenticated encrypted data. Content descriptor(s) (e.g., playlist(s)) can be modified to facilitate exchange of a token for a decryption key for browser(s) that do not provide logic to manage a flow of the token.

Abstract: A method includes obtaining playback experience data indicative of playback quality of a playback, determining a playback score for the playback based on the obtained playback experience data, analyzing the playback score to determine whether a playback issue is present, and providing a notification at the client indicating that the playback issue is present.

Abstract: Described herein are various technologies pertaining to delivery of token-authenticated encrypted data. Content descriptor(s) (e.g., playlist(s)) can be modified to facilitate exchange of a token for a decryption key for browser(s) that do not provide logic to manage a flow of the token.

Abstract: A multispeed playback system is described herein that allows for playback of smooth streaming media presentations at speeds other than the normal speed or direction, while still using an underlying platform that does not natively support multispeed. The system provides frames to the underlying platform so that the platform is still playing 1× content while the system manipulates video and/or audio samples being fed to the platform to make rendered data look appropriately for a selected speed and direction. The system may feed the underlying platform frames at a faster rate or skip every other frame for faster playback, or may feed each frame to the underlying platform multiple times for slower playback. Thus, the multispeed playback system allows a client application developer to provide users with expected multispeed playback at the client without encoding multiple different streams at the server.

Abstract: A streaming composition system is described herein that provides easy workflow and playback capabilities for content producers to create composite media assets from existing and on-going media content and for streaming clients to seamlessly playback composite multimedia streams provided from different sources. These assets provide broadcasters an option to quickly turn around highlights for an on-going event. The streaming composition system allows a producer to identify clips within existing media assets and compose the clips into a new unified streaming presentation. For producers that already have smooth streaming media assets, the system leverages these assets to provide seamless playback across clip boundaries including advanced playback support for advertisement insertion, fast forward, rewind, and so on.

Abstract: A streaming composition system is described herein that provides easy workflow and playback capabilities for content producers to create composite media assets from existing and on-going media content and for streaming clients to seamlessly playback composite multimedia streams provided from different sources. These assets provide broadcasters an option to quickly turn around highlights for an on-going event. The streaming composition system allows a producer to identify clips within existing media assets and compose the clips into a new unified streaming presentation. For producers that already have smooth streaming media assets, the system leverages these assets to provide seamless playback across clip boundaries including advanced playback support for advertisement insertion, fast forward, rewind, and so on.

Abstract: Described herein are various technologies pertaining to delivery of token-authenticated encrypted data. Content descriptor(s) (e.g., playlist(s)) can be modified to facilitate exchange of a token for a decryption key for browser(s) that do not provide logic to manage a flow of the token.

Abstract: A sparse streaming system provides a first-class means for sparse metadata to be added to streaming media presentations and to be delivered using an integrated data channel that is cacheable using readily available HTTP-based Internet caching infrastructure for increased scalability. The sparse streaming system stores a reference to a sparse track within a continuous track. If a continuous fragment arrives at the client that refers to a sparse fragment that the client has not yet retrieved, then the client requests the sparse fragment. In addition, each sparse fragment may include a backwards reference to the sparse fragment created immediately prior. The references in the continuous fragments make the client aware of new sparse track fragments, and the backwards references in the sparse track fragments ensure that the client has not missed any intervening sparse track fragments.