Vijay A Balasubramaniyan

US Patent:

20130109358, May 2, 2013

Filed:

Jun 29, 2011

Appl. No.:

13/807837

Inventors:

Vijay Balasubramaniyan - Atlanta GA, US
Mustaque Ahamad - Atlanta GA, US
Patrick Gerard Traynor - Decatur GA, US
Michael Thomas Hunter - Atlanta GA, US
Aamir Poonawalla - Atlanta GA, US

International Classification:

H04W 12/02

US Classification:

455411

Abstract:

Various embodiments of the invention are detection systems and methods for detecting call provenance based on call audio. An exemplary embodiment of the detection system can comprise a characterization unit, a labeling unit, and an identification unit. The characterization unit can extract various characteristics of networks through which a call traversed, based on call audio. The labeling unit can be trained on prior call data and can identify one or more codecs used to encode the call, based on the call audio. The identification unit can utilize the characteristics of traversed networks and the identified codecs, and based on this information, the identification unit can provide a provenance fingerprint for the call. Based on the call provenance fingerprint, the detection system can identify, verify, or provide forensic information about a call audio source.

US Patent:

20080256256, Oct 16, 2008

Filed:

Apr 10, 2007

Appl. No.:

11/733221

Inventors:

Arup Acharya - Nanuet NY, US
Vijay A. Balasubramaniyan - Atlanta GA, US
Mustaque Ahamad - Atlanta GA, US

Assignee:

International Business Machines Corporation - Armonk NY

International Classification:

G06F 15/173

US Classification:

709238

Abstract:

Systems and methods are provided for maximizing call throughput in a server network by optimizing the balance of stateful to stateless handling or transactions at each server within the network. The identification of transaction messages to be handled statelessly or statefully is made at each proxy server within the network in order to maximize the total throughput at that proxy server within prescribed processor utilization limits. In general, each transaction is handled statefully by at least one server within the network. Reports on the stateful handling of messages and the resource consumption at various proxies are communicated throughout the network to be used in identifying the ratio of messages to be forwarded statefully to messages to be forwarded statelessly at any given proxy.

US Patent:

20220392452, Dec 8, 2022

Filed:

Jun 3, 2022

Appl. No.:

17/832146

Inventors:

- Atlanta GA, US
Elie KHOURY - Atlanta GA, US
Terry NELMS, II - Atlanta GA, US
Vijay BALASUBRAMANIYAN - Atlanta GA, US

Assignee:

Pindrop Security, Inc. - Atlanta GA

International Classification:

G10L 17/04
G06F 21/32
G10L 17/22

Abstract:

Disclosed are systems and methods including computing-processes executing machine-learning architectures extract vectors representing disparate types of data and output predicted identities of users accessing computing services, without express identity assertions, and across multiple computing services, analyzing data from multiple modalities, for various user devices, and agnostic to architectures hosting the disparate computing service. The system invokes the identification operations of the machine-learning architecture, which extracts biometric embeddings from biometric data and context embeddings representing all or most of the types of metadata features analyzed by the system. The context embeddings help identify a subset of potentially matching identities of possible users, which limits the number of biometric-prints the system compares against an inbound biometric embedding for authentication. The types of extracted features originate from multiple modalities, including metadata from data communications, audio signals, and images. In this way, the embodiments apply a multi-modality machine-learning architecture.

US Patent:

20220392453, Dec 8, 2022

Filed:

Jun 3, 2022

Appl. No.:

17/832404

Inventors:

- Atlanta GA, US
Elie KHOURY - Atlanta GA, US
Terry Nelms, II - Atlanta GA, US
Vijay BALASUBRAMANIYAN - Atlanta GA, US

Assignee:

Pindrop Security, Inc. - Atlanta GA

International Classification:

G10L 17/04
G10L 17/12
G06F 21/32

Abstract:

Disclosed are systems and methods including computing-processes executing machine-learning architectures extract vectors representing disparate types of data and output predicted identities of users accessing computing services, without express identity assertions, and across multiple computing services, analyzing data from multiple modalities, for various user devices, and agnostic to architectures hosting the disparate computing service. The system invokes the identification operations of the machine-learning architecture, which extracts biometric embeddings from biometric data and context embeddings representing all or most of the types of metadata features analyzed by the system. The context embeddings help identify a subset of potentially matching identities of possible users, which limits the number of biometric-prints the system compares against an inbound biometric embedding for authentication. The types of extracted features originate from multiple modalities, including metadata from data communications, audio signals, and images. In this way, the embodiments apply a multi-modality machine-learning architecture.

US Patent:

20210150010, May 20, 2021

Filed:

Jan 25, 2021

Appl. No.:

17/157837

Inventors:

- Atlanta GA, US
Kailash Patil - Atlanta GA, US
David Dewey - Atlanta GA, US
Raj Bandyopadhyay - Atlanta GA, US
Telvis Calhoun - Atlanta GA, US
Vijay Balasubramaniyan - Atlanta GA, US

International Classification:

G06F 21/32
G06N 20/00
G06F 21/55
H04M 3/493
H04W 12/128
H04M 3/527
H04M 15/00
H04W 12/12

Abstract:

Systems and methods for call detail record (CDR) analysis to determine a risk score for a call and identify fraudulent activity and for fraud detection in Interactive Voice Response (IVR) systems. An example method may store information extracted from received calls. Queries of the stored information may be performed to select data using keys, wherein each key relates to one of the received calls, and wherein the queries are parallelized. The selected data may be transformed into feature vectors, wherein each feature vector relates to one of the received calls and includes a velocity feature and at least one of a behavior feature or a reputation feature. A risk score for the call may be generated during the call based on the feature vectors.

US Patent:

20190342452, Nov 7, 2019

Filed:

Jul 18, 2019

Appl. No.:

16/515823

Inventors:

- Atlanta GA, US
Kailash PATIL - Atlanta GA, US
David DEWEY - Atlanta GA, US
Raj BANDYOPADHYAY - Atlanta GA, US
Telvis CALHOUN - Atlanta GA, US
Vijay BALASUBRAMANIYAN - Atlanta GA, US

International Classification:

H04M 3/527
H04M 15/00
G06F 21/55
H04M 3/493
G06F 21/32
G06N 20/00

Abstract:

Systems and methods for call detail record (CDR) analysis to determine a risk score for a call and identify fraudulent activity and for fraud detection in Interactive Voice Response (IVR) systems. An example method may store information extracted from received calls. Queries of the stored information may be performed to select data using keys, wherein each key relates to one of the received calls, and wherein the queries are parallelized. The selected data may be transformed into feature vectors, wherein each feature vector relates to one of the received calls and includes a velocity feature and at least one of a behavior feature or a reputation feature. A risk score for the call may be generated during the call based on the feature vectors.

US Patent:

20180152561, May 31, 2018

Filed:

Jan 25, 2018

Appl. No.:

15/880287

Inventors:

- Atlanta GA, US
Kailash PATIL - Atlanta GA, US
David DEWEY - Atlanta GA, US
Raj BANDYOPADHYAY - Atlanta GA, US
Telvis CALHOUN - Canton GA, US
Vijay BALASUBRAMANIYAN - Atlanta GA, US

Assignee:

PINDROP SECURITY, INC. - Atlanta GA

International Classification:

H04M 3/527
G06N 99/00
H04M 15/00
H04M 7/00
H04W 12/12

Abstract:

Systems and methods for call detail record (CDR) analysis to determine a risk score for a call and identify fraudulent activity and for fraud detection in Interactive Voice Response (IVR) systems. An example method may store information extracted from received calls. Queries of the stored information may be performed to select data using keys, wherein each key relates to one of the received calls, and wherein the queries are parallelized. The selected data may be transformed into feature vectors, wherein each feature vector relates to one of the received calls and includes a velocity feature and at least one of a behavior feature or a reputation feature. A risk score for the call may be generated during the call based on the feature vectors.

US Patent:

20170111506, Apr 20, 2017

Filed:

Oct 14, 2016

Appl. No.:

15/294538

Inventors:

- Atlanta GA, US
Kailash PATIL - Atlanta GA, US
David DEWEY - Atlanta GA, US
Raj BANDYOPADHYAY - Atlanta GA, US
Telvis CALHOUN - Canton GA, US
Vijay BALASUBRAMANIYAN - Atlanta GA, US

Assignee:

PINDROP SECURITY, INC. - Atlanta GA

International Classification:

H04M 3/527
G06N 99/00
H04M 15/00

Abstract:

Systems and methods for call detail record (CDR) analysis to determine a risk score for a call and identify fraudulent activity and for fraud detection in Interactive Voice Response (IVR) systems. An example method may store information extracted from received calls. Queries of the stored information may be performed to select data using keys, wherein each key relates to one of the received calls, and wherein the queries are parallelized. The selected data may be transformed into feature vectors, wherein each feature vector relates to one of the received calls and includes a velocity feature and at least one of a behavior feature or a reputation feature. A risk score for the call may be generated during the call based on the feature vectors.