Autonomous configuration modeling and management

1. A method, comprising:

probing at least one element of a set of elements of a networked architecture, the networked architecture comprising software elements and hardware elements interconnected in a common environment, and wherein the probing compiles information about the at least one element of the set of elements;

learning a first configuration for the at least one element in the common environment based on the probing, wherein the learning comprises:

determining the set of elements of the networked architecture, the set of elements having identifiers;

determining a plurality of types of data sources associated with the set of elements using the identifiers, wherein the plurality of types of data sources comprise external data sources having information about the set of elements, the plurality of types of data sources comprising behavior data describing performance of the set of elements or interactions with the set of elements; and

compiling data from the plurality of types of data sources according to a learned priority of the plurality of types of data sources, the learned priority indicating the most relevant data sources to the set of elements learned over time;

determining at least one vulnerability in the learned first configuration;

analyzing the at least one vulnerability;

determining a threat score for the at least one vulnerability based on the analysis;

developing a threat model based on the learned first configuration and the threat score for the at least one vulnerability;

applying the threat model to the networked architecture to determine a second configuration that resolves the at least one vulnerability; and

deploying the second configuration wherein the deployment of the second configuration is prioritized based on the threat score.

2. The method of claim 1 , wherein the second configuration includes at least one of an installation, an upgrade, a patch, or an uninstallation.

3. The method of claim 1 , wherein analyzing the vulnerability comprises: applying deep learning to the vulnerability using at least one learned data source; and compiling the threat score based on the deep learning of the vulnerability.

4. The method of claim 1 , wherein the plurality of types of data sources comprise at least one of system behavior, human behavior, previous configurations, exploitation signatures, failed configurations, internet database, or intranet database.

5. The method of claim 1 , comprising:

compiling the data associated with the set of elements into a knowledgebase; and

determining via machine learning the data of the knowledgebase; and

continuously updating the threat model according to the machine learning of the information.

6. A system, comprising:

one or more processors;

a non-transitory computer-readable storage medium having program code that is executable by the one or more processors to cause the one or more processors to perform operations comprising:

probing at least one element of a set of elements of a networked architecture, the networked architecture comprising software elements and hardware elements interconnected in a common environment, and wherein the probing compiles information about the at least one element of the set of elements; learning a first configuration for the at least one element in the common environment based on the probing, wherein the learning comprises: determining the set of elements of the networked architecture, the set of elements having identifiers; determining a plurality of types of data sources associated with the set of elements using the identifiers, wherein the plurality of types of data sources comprise external data sources having information about the set of elements, the plurality of types of data sources comprising behavior data describing performance of the set of elements or interactions with the set of elements; and compiling data from the plurality of types of data sources according to a learned priority of the plurality of types of data sources, the learned priority indicating the most relevant data sources to the set of elements learned over time; determining at least one vulnerability in the learned first configuration; analyzing the at least one vulnerability; determining a threat score for the determined vulnerability and the at least one element developing a threat model based on the learned first configuration and the threat score for the at least one vulnerability; and applying the threat model to the networked architecture to determine a second configuration that resolves the at least one vulnerability; and deploying the second configuration wherein the deployment of the second configuration is prioritized based on the threat score.

7. The system of claim 6 , wherein the second configuration includes at least one of an installation, an upgrade, a patch, or uninstallation.

8. The system of claim 6 , wherein the operations further comprise:

applying deep learning to the vulnerability using at least one learned data source; and

compiling the threat score based on the deep learning of the vulnerability.

9. The system of claim 6 , wherein the plurality of types of data sources comprise at least one of system behavior, human behavior, previous configurations, exploitation signatures, failed configurations, internet database, or intranet database.

10. The system of claim 6 , wherein the operations further comprise:

compiling data associated with the set of elements into a knowledgebase; and

learning the data of the knowledgebase using machine learning; and

continuously updating the threat model according to the machine learning of the data.

11. A non-transitory computer readable medium having instructions executable by one or more processors to cause the one or more processor to:

probe at least one element of a set of elements of a networked architecture, the networked architecture comprising software elements and hardware elements interconnected in a common environment, and wherein the probing compiles information about the at least one element of the set of elements;

learn a first configuration for the at least one element in the common environment based on the probing, wherein the learning comprises:

determining the set of elements of the networked architecture, the set of elements having identifiers;

determining a plurality of types of data sources associated with the set of elements using the identifiers, wherein the plurality of types of data sources comprise external data sources having information about the set of elements, the plurality of types of data sources comprising behavior data describing performance of the set of elements or interactions with the set of elements; and

compiling data from the plurality of types of data sources according to a learned priority of the plurality of types of data sources, the learned priority indicating the most relevant data sources to the set of elements learned over time;

determine at least one vulnerability in the learned first configuration;

analyze the at least one vulnerability;

determine a threat score for the determined vulnerability and the at least one element;

develop a threat model based on the learned first configuration and the threat score for the at least one vulnerability;

apply the threat model to the networked architecture to determine a second configuration that resolves the at least one vulnerability; and

deploy the second configuration based on the threat model to the set of elements in the networked architecture, wherein the deployment of the second configuration is prioritized based on the threat score.

12. The non-transitory computer readable medium of claim 11 , wherein the instructions further cause the one or more processors to:

learn data of the set of elements and vulnerability using deep learning; and

continuously update the threat model according to the learning of the data.