Abstract: An illustrative method includes monitoring activities within a compute environment and generating a logical graph model using at least a portion of the monitored activities. The logical graph model includes a set of nodes representative of one or more identities, applications, and resources in the compute environment and a set of edges representative of connections between nodes interconnected by the edges. An operation associated with security of the compute environment is performed using the generated logical graph model.

Abstract: Systems and methods are provided for enabling adaptive modifications to AI agents using human feedback, particularly in the context of investigating cybersecurity alerts. According to one implementation, a method includes a step of receiving feedback from a human analyst related to results of a task performed by an Artificial Intelligence (AI) agent. The method can include a step of converting the feedback into a structured representation having nodes and edges. Furthermore, the method includes a step of updating a knowledge database associated with the AI agent using the structured representation. Next, the method includes a step of utilizing the structured representation and/or knowledge database to improve performance of the AI agent with respect to subsequent tasks.

Abstract: Systems and methods for automatically investigating potential cyber security alerts threats are provided. A method includes receiving logs related to security alerts from multiple sources, the security alerts representing potential cyber security threats in a compute domain; and performing an automated investigation procedure configured to determine whether the logs represent actual cyber security threats, the automated investigation procedure including a plan generation stage in which high-level logical steps are planned for analyzing the logs and retrieving evidences for proving it either malicious or benign, a log comprehension stage in which details of the logs are analyzed to obtain observations of the logs, a plan execution stage in which the steps of the plan generation stage are executed with respect to the observations of the logs, a reasoning stage to conclude the case, and a re-planning stage to generate a new investigation plan for newly discovered entities or signal.

Abstract: An illustrative method for performing a risk scenario assessment and remediation may include identifying, based on posture data associated with a compute environment, one or more compute resources deployed in the compute environment that are configured to be connected to a network, accessing runtime workload data associated with the one or more compute resources representative of network activity for the one or more compute resources, and performing, based on the posture data and the runtime workload data, a remediation operation associated with the one or more compute resources.

Abstract: An illustrative method for performing an identification of secrets misuse may include accessing runtime workload data associated with one or more keys configured to be accessed by a workload deployed in a compute environment, detecting, based on the runtime workload data, an anomaly associated with the one or more keys, and determining, based on the anomaly, that secrets misuse has possibly occurred within the compute environment.

Abstract: An illustrative method includes identifying, based on a scan of a compute environment associated with an entity, a plurality of attack paths from one or more networks to one or more datasets associated with the entity and determining a set of one or more attack paths included in the plurality of attack paths that include a particular risk artifact. Based on one or more characteristics of the set of one or more attack paths, a risk score specific to the particular risk artifact may be determined and a risk mitigation operation associated with the particular risk artifact may be performed.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating machine code for a quantized neural network. In one aspect, one of the methods include: receiving source code for a neural network, and performing a profile-guided quantization process to generate an optimized quantization configuration for the neural network that defines, for each of the plurality of layers, one or more optimized computer number formats for representing activation values generated by the layer, the respective sets of parameters for the layer, or both.

Abstract: An illustrative method includes identifying, based on static workload data associated with a compute environment, one or more attack paths from a network to one or more datasets associated with an entity, accessing runtime workload data associated with the compute environment, and performing, based on the runtime workload data, a risk mitigation operation associated with the one or more attack paths.

Abstract: The present disclosure discloses a water outlet mechanism comprising a water outlet body, a driving assembly, an eccentric cam, a swinging plate, and one or more water outlet nozzles. The water outlet body comprises a water inlet port, a water outlet port, and a water-passing cavity in communication with the water inlet port and the water outlet port. The driving assembly is disposed in the water-passing cavity. The eccentric cam is rotatably disposed in the water-passing cavity and is operatively coupled to the driving assembly. The swinging plate is swingingly disposed in the water-passing cavity and is operatively coupled to the eccentric cam. The one or more water outlet nozzles are swingingly disposed on the swinging plate and are configured to be driven by the swinging plate to perform reciprocating movement of swinging in a front-and-rear direction.

Abstract: An illustrative method includes scanning a compute environment associated with an entity and identifying one or more attack paths from a network to one or more datasets associated with the entity. The one or more attack paths each include a series of risk artifacts within the compute environment that can be exploited by an attacker to access the one or more datasets. The method further includes generating one or more attack path risk scores associated with the one or more attack paths and indicative of one or more levels of risk that the one or more attack paths could be exploited to access the one or more datasets. A risk mitigation operation associated with the one or more attack paths is performed based on the one or more attack path risk scores.

Abstract: An illustrative method includes accessing, by a data platform configured to monitor a compute environment, permissions data representative of a set of permissions that specify how an entity is entitled to interact with resources within the compute environment; collecting, by the data platform, runtime workload data representative of an actual interaction by the entity with the resources over a period of time; and performing, by the data platform based on the runtime workload data, an operation associated with modifying the set of permissions for the entity.

Abstract: An illustrative method includes tracking, by a data platform configured to monitor a compute environment, a plurality of identity transitions that occur over time with respect to an entity, wherein each of the identity transitions includes a transition by the entity from being associated with one identity to being associated with another identity, the one identity and the another identity having different permission sets with respect to resources within the compute environment; determining, by the data platform while performing the tracking, that an attribute of the plurality of identity transitions satisfies a predetermined criterion; and performing, by the data platform based on the attribute of the plurality of transitions satisfying the predetermined criterion, a remedial action associated with the entity.

Abstract: Using natural language queries to conduct an investigation of a monitored system, the method comprising: receiving a natural language input; generating, by a large language model and based on the natural language input, a first query directed to one or more tables of a plurality of tables; generating, based on the first query, one or more second queries each directed to a different table of the one or more tables; and presenting, via a user interface, a response to each of the one or more second queries.

Abstract: An illustrative method for operating an agent configuration may include providing an unprivileged agentless workload scanning configuration to collect non-runtime workload data associated with a workload deployed within a cloud environment, determining one or more properties of the workload based on the non-runtime workload data, and performing an operation with respect to a privileged agent configuration deployed within the cloud environment and configured to collect runtime workload data associated with the workload.