Patents by Inventor Kamran Razi

Kamran Razi has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Customized anomaly detection in sandbox software security systems using graph convolutional networks
    Patent number: 12197562
    Abstract: An ML (machine learning) training logs are parsed for generating a set of heterogenous graphs having embedded nodes connected with edges determined with link prediction and denoting a hierarchical relationship between nodes. Each graph represents benign behavior from executing one of the files of a training database in the sandbox, wherein the nodes are embedded in the graph using GCN (graph convolution network) to calculate a real-valued vector with fixed dimension. A runtime module to receive an untagged file in real-time for analysis from a network component, and generates a graph of runtime behavior from sandbox of the suspicious file for comparison against the training graphs.
    Type: Grant
    Filed: December 31, 2021
    Date of Patent: January 14, 2025
    Assignee: Fortinet, Inc.
    Inventors: Kamran Razi, Jun Cai
  • Early malware detection in on-the-fly security sandboxes using recursive neural networks (RNNs)to capture relationships in behavior sequences on data communication networks
    Patent number: 11968228
    Abstract: A file copy is executed in a virtual runtime environment that tracks behavior using RNN taking runtime behavior of at least a first time into account with current runtime behavior at a second time. This is responsive to not finding a known signature for suspicious activity during virus scanning. A behavior sequence is identified on-the-fly during file copy execution that is indicative of malware, prior to completing the execution, the behavior sequence involving at least two actions taken at different times during file copy execution. Responsive to the identification, the execution is terminated and the virtual runtime environment is returned to the pool of available virtual runtime environments.
    Type: Grant
    Filed: December 9, 2020
    Date of Patent: April 23, 2024
    Assignee: Fortinet, Inc.
    Inventors: Jun Cai, Kamran Razi
  • CUSTOMIZED ANOMALY DETECTION IN SANDBOX SOFTWARE SECURITY SYSTEMS USING GRAPH CONVOLUTIONAL NETWORKS
    Publication number: 20230222208
    Abstract: An ML (machine learning) training logs are parsed for generating a set of heterogenous graphs having embedded nodes connected with edges determined with link prediction and denoting a hierarchical relationship between nodes. Each graph represents benign behavior from executing one of the files of a training database in the sandbox, wherein the nodes are embedded in the graph using GCN (graph convolution network) to calculate a real-valued vector with fixed dimension. A runtime module to receive an untagged file in real-time for analysis from a network component, and generates a graph of runtime behavior from sandbox of the suspicious file for comparison against the training graphs.
    Type: Application
    Filed: December 31, 2021
    Publication date: July 13, 2023
    Inventors: Kamran Razi, Jun Cai
  • EARLY MALWARE DETECTION IN ON-THE-FLY SECURITY SANDBOXES USING RECURSIVE NEURAL NETWORKS (RNNS)TO CAPTURE RELATIONSHIPS IN BEHAVIOR SEQUENCES ON DATA COMMUNICATION NETWORKS
    Publication number: 20220182395
    Abstract: A file copy is executed in a virtual runtime environment that tracks behavior using RNN taking runtime behavior of at least a first time into account with current runtime behavior at a second time. This is responsive to not finding a known signature for suspicious activity during virus scanning. A behavior sequence is identified on-the-fly during file copy execution that is indicative of malware, prior to completing the execution, the behavior sequence involving at least two actions taken at different times during file copy execution. Responsive to the identification, the execution is terminated and the virtual runtime environment is returned to the pool of available virtual runtime environments.
    Type: Application
    Filed: December 9, 2020
    Publication date: June 9, 2022
    Inventors: Jun Cai, Kamran Razi
  • ADAPTIVE FILTERING OF MALWARE USING MACHINE-LEARNING BASED CLASSIFICATION AND SANDBOXING
    Publication number: 20220067146
    Abstract: Systems and methods for adaptive filtering of malware using a machine-learning model and sandboxing are provided. According to one embodiment, a processing resource of a sandbox appliance receives a file. A feature vector associated with the file is generated by extracting multiple static features from the file. The file is classified based on the feature vector by applying a machine-learning model. When the classification of the file is unknown, representing insufficient information is available to identify the file as malicious or benign, sandbox processing is caused to be performed on the file.
    Type: Application
    Filed: September 1, 2020
    Publication date: March 3, 2022
    Applicant: Fortinet, Inc.
    Inventors: Jun Cai, Kamran Razi