Abstract: Described herein are various examples of techniques for evaluation of entities communicating over a network. An example method may include receiving a set of network traffic data and a potential bot candidate associated with the set of network traffic data. The example method may also include adjusting a bot risk assessment associated with the potential bot candidate by analyzing a subset of the set of network traffic data corresponding to the potential bot candidate via an expert machine learning model trained to detect attack vectors based on network traffic data associated with potential bot candidates. The example method may further include outputting the bot risk assessment associated with the potential bot candidate. Various other systems, methods, and computer-readable media are also disclosed.

Abstract: Described herein are various examples of techniques for generating a network configuration and configuring a network based on said configuration. In some embodiments, there is provided a method comprising receiving a natural language input describing a desired behavior of a network and generating a network configuration based at least in part in the natural language input. Generating the network configuration may comprise generating structured data indicating one or more rules for the network to implement the desired behavior. The method may further comprise configuring the network based on the structured data indicating the one or more rules to implement the desired behavior of the network.

Abstract: Disclosed are systems and methods for providing unified routing and caching logic for applications utilizing a CDN. First, the system maintains access to one or more edge servers for a CDN, with the one or more edge servers being configured to execute edge logic. The system receives code for an application in a specified application framework, with the code and/or application framework including a unified set of routing and caching logic for dynamic data to be applied to a number of caches, and with the unified set of routing and caching logic being derived from a centralized definition. For each of the caches, the system first analyzes the set of routing and caching logic to determine a subset of the routing and caching logic to be executed by the cache, and then atomically sends the determined subset of the routing and caching logic to the cache.

Abstract: Described herein are various examples of techniques for server-side identification of an entity communicating over a network, which may in some embodiments include techniques for identifying entities communicating in a network based on a signature for the entity and/or behavior of the entity as determined or observed by one or more servers in the network. In some embodiments, server-side identification may include signature analysis, including collecting information regarding the entity from communications received at the server and/or received at an intermediary server that may perform caching functionality and using such information to determine a signature of the entity.

Abstract: Disclosed are systems and methods for enabling an application developer to define CDN behavior (e.g., caching rules) alongside an application program code and embedded in the same source code. The systems and methods can be used to define CDN or edge caching behavior using custom cache keys that go beyond simply using a URL. The custom cache keys and source code CDN behavior definitions can be used to enable a graphical user interface to specify traffic rules in a codeless interface. The described embodiments can be used for defining general traffic routing, to test different versions of the application or to otherwise present different versions of the application to end users, while preserving cache keys and caching behavior at the edge and throughout the application distribution layers.

Abstract: Disclosed are systems and methods for providing automated configuration of a content delivery network. First, the system maintains access to one or more edge servers for a CDN, with the edge servers being configured to execute edge logic. The system receives code for an application in a specified application framework recognized by the system, where the code includes a set of routing and caching logic with routing instructions and routing behaviors. The system then analyzes the set of routing and caching logic to define a set of edge routing behaviors and caching behaviors for one or more routes within the application. The system then determines at least a subset of the routing and caching behaviors that can be processed at the one or more edge servers using the edge logic.

Abstract: Disclosed is an activity detection system (“ADS”) that detects, classifies, and isolates previously unseen activities in unlabeled videos based on different previously seen activities in labeled videos and semantic similarity between the unseen and seen activities. The ADS receive a first set of videos that are labeled with a first activity, and may determine a feature set within frames of the first set of videos that represents the first activity. The ADS may receive a second set of videos that are not labeled, and a query for videos of a second activity that is determined to be semantically similar to the first activity. The ADS may provide, in response to the query for the second activity, a particular video from the second set of videos containing the feature set representing the first activity that is semantically similar to the queried for second activity.

Abstract: Intelligent dynamic preloading is described. A list of page requests is generated using a prediction model that predicts a plurality of page requests for a site or web application that are expected to be received from client devices, where at least some of the page requests include URLs and one or more other parameters that affect generation of a corresponding page. A request is simulated for each of the page requests included on the list of page requests to a caching layer to cause the caching layer to: make a request to an origin server for the site or dynamic web application, receive rendered pages from the origin server, and store the received rendered pages, such that subsequent requests received at the caching layer can be served from the already rendered pages.

Abstract: Provided is a Routing Early Warning System (“REWS”) that preemptively detects and corrects network issues based on control plane messaging. REWS receives control plane messages for network paths to a source node, groups the control plane messages to different bins based on time, detects an anomaly based on a number of a first set of control plane messages grouped to a particular bin differing, by a threshold amount, from a steady state number of control plane messages grouped to at least one other bin, and isolates a cause of the anomaly based on a number of updated paths and addressing of one or more nodes specified in the first set of control plane messages. REWS modifies routing of the source node data plane traffic before the anomaly significant impacts the data plane in response to detecting the anomaly and isolating the anomaly cause using the control plane messages.

Abstract: Dynamically transforming web content is described. An HTTP request is received from an Internet client. The web resource identified in the HTTP request is accessed. The content of the web resource is analyzed. A set of transformation instructions are applied on a set of identified portions of the content of the web resource. Each applied transformation instruction includes logic to locate and manipulate at least an identified portion of the content, and at least one of the applied transformation instructions is a client-side script transformation instruction that performs one or more of: modify a client-side script included in the content, remove a client-side script included in the content, and add a client-side script to the content. An HTTP response is rendered that includes the results of the applied transformation instructions and further includes those portions of the content that were not manipulated by a transformation instruction. The response is then transmitted to the Internet client.

Abstract: Dynamically transforming web content is described. An HTTP request is received from an Internet client. The web resource identified in the HTTP request is accessed. The content of the web resource is analyzed. A set of transformation instructions are applied on a set of identified portions of the content of the web resource. Each applied transformation instruction includes logic to locate and manipulate at least an identified portion of the content, and at least one of the applied transformation instructions is a client-side script transformation instruction that performs one or more of: modify a client-side script included in the content, remove a client-side script included in the content, and add a client-side script to the content. An HTTP response is rendered that includes the results of the applied transformation instructions and further includes those portions of the content that were not manipulated by a transformation instruction. The response is then transmitted to the Internet client.

Abstract: Provided is an unobtrusive client verification system with one verification devices having processors that are configured to receive a first request from an unverified client device, generate a random number in response to receiving the first request from the unverified client device, define a set of expressions as a browser challenge problem that evaluates to an answer specified by the random number, encrypt the answer within an answer token, provide the browser challenge problem with the answer token to the unverified client device, receive a second request with a solution to the browser challenge problem and the answer token from the unverified client device, and verify the unverified client device in response to the solution matching the answer that is decrypted from the answer token provided with the second request.

Abstract: Disclosed are systems and methods for enabling an application developer to define CDN behavior (e.g., caching rules) alongside an application program code and embedded in the same source code. The systems and methods can be used to define CDN or edge caching behavior using custom cache keys that go beyond simply using a URL. The custom cache keys and source code CDN behavior definitions can be used to enable a graphical user interface to specify traffic rules in a codeless interface. The described embodiments can be used for defining general traffic routing, to test different versions of the application or to otherwise present different versions of the application to end users, while preserving cache keys and caching behavior at the edge and throughout the application distribution layers.

Abstract: Provided is an Adaptive Network Security System (“ANSS”) that receives a first set of network data, detects commonality in a set of parameters within the first set of network data using an unsupervised machine learning technique, and models an expected behavior based on the commonality in the set of parameters. The ANSS may determine a threat risk associated with a second set of network data based on an amount of deviation between the set of parameters from the second plurality of network data and the expected behavior, and may perform a particular action from different actions against the second set of network data in response to the set of parameters from the second set of network data deviating from the expected behavior and the threat risk.

Abstract: Provided is a Routing Early Warning System (“REWS”) that preemptively detects and corrects network issues based on control plane messaging. REWS receives control plane messages for network paths to a source node, groups the control plane messages to different bins based on time, detects an anomaly based on a number of a first set of control plane messages grouped to a particular bin differing, by a threshold amount, from a steady state number of control plane messages grouped to at least one other bin, and isolates a cause of the anomaly based on a number of updated paths and addressing of one or more nodes specified in the first set of control plane messages. REWS modifies routing of the source node data plane traffic before the anomaly significant impacts the data plane in response to detecting the anomaly and isolating the anomaly cause using the control plane messages.