Kyndryl, Inc.

International Business Machines Corporation

Sreekrishnan Venkateswaran

Manish Satyapal Gupta

Ashish Kundu

Anuradha Bhamidipaty

Debasisha Padhi

Deepak Poola Chandrashekar

Muhammad Ammar Ahmed

Madiha Ijaz

Shiva P Chandrashekar

Ashoka S Rao

Vugranam Chakravarthy Sreedhar

Charanjit Singh Jutla

Sridhar Rao

Shubhi Asthana

Seeta Hariharan

Jigar Navanitrai Kapasi

Nicholas M Williamson

Brooks Johnston

Bahram Sanaei

Raghava Reddy Kapu Veera

Madhusudan Kunigal Satyanarayana

Poornima Sathyanaraya

Santosh Thomas

Piotr Uzar

A method and related apparatus adaptively control snapshot replication of a plurality of server snapshots in a multi-tenant public cloud using snapshot service. A plurality of snapshot requests are received from a plurality of clients in the multi-tenant public cloud and are each associated with a service level agreement (SLA). The SLA includes a corresponding completion deadline and a slippage penalty. A probabilistic model calculates, for active snapshot flows, a completion time for each of the snapshot requests. If any of the predicted completion times exceed their corresponding completion deadlines, a possible MISS event is produced for an associated first snapshot. The snapshot controller component, in response to producing the possible MISS event, minimizes a probability of breaching fulfilment times for the multi-tenant cloud. The slippage penalty is calculated for each of the plurality of snapshot requests. A second snapshot is identified from among the plurality of snapshot requests.

Adaptive snapshot controller

An example operation may include one or more of capturing a current version of sensitive data by a data processor node, hashing, by the data processor node, the current version of the sensitive data, storing, by the data processor node, a hash of the current version of the sensitive data on a first blockchain, encrypting, by the data processor node, the current version of the sensitive data using a secret key, and storing the encrypted current version of the sensitive data on a second blockchain.

Protecting sensitive data

An embodiment extracts a set of designated entities and a set of relationships between designated entities from speech content of an audio feed of a plurality of participants of a current web conference using a machine learning model trained to classify parts of speech content. The embodiment generates a list of current action items based on the extracted set of designated entities and relationships between designated entities. The embodiment identifies a first current action item that is an updated version of an ongoing action item on a progress list of ongoing action items from past web conferences. The embodiment also identifies a second current action item that is unrelated to any of the ongoing action items on the progress list. The embodiment updates the progress list to include updates for the first current action item and by adding the second current action item.

Cognitive correlation of group interactions

A computer-implemented method in accordance with one embodiment includes, in response to a submission of an input dataset to an artificially intelligent application, receiving an explanation from each module of the application. The modules are configured within the application in a serial sequence in which each module, upon receiving the input dataset and any input generated by an immediately preceding module of the serial sequence, generates output that is forwarded as input to a next module, if any, in the sequence. A determination is made that at least two of the received explanations are semantically inconsistent.

Semantic consistency of explanations in explainable artificial intelligence applications

An embodiment includes identifying which of a plurality of participants of a web conference is an identified participant associated with a selected cluster of a plurality of clusters of audio feed data of an audio feed of the web conference based on a self-introduction in the selected cluster. The embodiment also generates a first preliminary leadership score for the identified participant based on a speaking duration value associated with the identified participant and generates a second preliminary leadership score for the identified participant using a selected video segment as an input for a machine learning classifier model. The embodiment calculates a final leadership score for the identified participant based on the first and second preliminary leadership scores. The final leadership score is representative of a likelihood that the identified participant is a supervisor, and is indicative of the identified participant being a supervisor if it exceeds a designated threshold value.

Cognitive encapsulation of group meetings

A method and related apparatus adaptively control snapshot replication of a plurality of server snapshots in a multi-tenant public cloud using snapshot service. A plurality of snapshot requests are received from a plurality of clients in the multi-tenant public cloud and are each associated with a service level agreement (SLA). The SLA includes a corresponding completion deadline and a slippage penalty. A probabilistic model calculates, for active snapshot flows, a completion time for each of the snapshot requests. If any of the predicted completion times exceed their corresponding completion deadlines, a possible MISS event is produced for an associated first snapshot. The snapshot controller component, in response to producing the possible MISS event, minimizes a probability of breaching fulfillment times for the multi-tenant cloud. The slippage penalty is calculated for each of the plurality of snapshot requests. A second snapshot is identified from among the plurality of snapshot requests.

Machine-based video classifying to identify misleading videos by training a model using a video corpus, obtaining a subject video from a content server, generating respective feature vectors of a title, a thumbnail, a description, and a content of the subject video, determining a first semantic similarities between ones of the feature vectors, determining a second semantic similarity between the title of subject video and titles of videos in the misleading video corpus in a same domain as the subject video, determining a third semantic similarity between comments of the subject video and comments of videos in the misleading video corpus in the same domain as the subject video, classifying the subject video using the model and based on the first semantic similarities, the second semantic similarity, and the third semantic similarity, and outputting the classification of the subject video to a user.

Automatic identification of misleading videos using a computer network

An explainable artificially intelligent (XAI) application contains an ordered sequence of artificially intelligent software modules. When an input dataset is submitted to the application, each module generates an output dataset and an explanation that represents, as a set of Boolean expressions, reasoning by which each output element was chosen. If any pair of explanations are determined to be semantically inconsistent, and if this determination is confirmed by further determining that an apparent inconsistency was not a correct response to an unexpected characteristic of the input dataset, nonzero inconsistency scores are assigned to inconsistent elements of the pair of explanations. If the application's overall inconsistency score exceeds a threshold value, the system forwards information about the explanation, the offending modules, and the input dataset to a downstream machine-learning component that uses this information to train the application to better respond to future input that shares certain characteristics with the current input.

A method, computer system, and a computer program product for multi-cloud solution design and fulfillment via crowdsourcing is provided. Embodiments of the present invention may include receiving a request, wherein the request includes computing requirements. Embodiments of the present invention may include parsing the received request into partitions. Embodiments of the present invention may include advertising the partitions to a plurality of participating members on a trusted network environment. Embodiments of the present invention may include receiving conceptual solutions to the partitions. Embodiments of the present invention may include evaluating the received conceptual solutions.

Automonous multi-cloud solution design and fulfillment via crowdsourcing

Metering based on application code complexity is provided. Code of an application to be executed is received from a user. Values of parameters corresponding to the application are extracted based on analysis of the code. A set of computational vectors corresponding to extracted values of the parameters of the application is generated to calculate a complexity of the code. A metering metric to charge the user to execute the application is determined based on the complexity of the code. The application is executed at the metering metric based on the complexity of the code.

Metering based on application code complexity

A computer implemented method for migrating streaming applications from on-premise non-cloud infrastructure to target clouds. The method may include receiving at the message connector a source message from an on-premise application. The method may further include analyzing the source message to assess feasibility of migrating the on-premise application to the target cloud wherein analysis includes extracting communication data from the source message and meta data from the source message. The method may also include converting the source message to a format native to the target cloud application, wherein the conversion by the message connector includes a cognitive engine to match a selected target cloud to the on premise application.

Message connector as a service to migrate streaming applications into cloud nativity

Aspects utilize a computing processing capability of a device connected to a cellular network wherein processors are configured to determine processing capabilities of each of a plurality of devices as function of device hardware configuration, software configuration, and average idle utilization, determine first remaining uptime periods of availability that each of the plurality of devices are available for data processing as a function of respective device data comprising battery level, signal strength and usage patterns, wherein the usage patterns are power usage patterns or data usage patterns, in response to receiving a processing task that comprises a needed processing capability, identify a subset of devices that each have a threshold amount of uptime of processing capability in an amount inclusive of the needed processing capability of the task, and assign the processing task to one of the subset devices.

Cellular hypervisor

A model drive system models tier changes within enterprise workloads by analyzing past data to automatically generate the capability to detect factors or situations which demand a change of service tier in a preferably hybrid cloud context with potentially multiple providers.

Model driven dynamic management of enterprise workloads through adaptive tiering

A system, method and program product for providing cognitive behavior security control (CBSC). A system is disclosed that includes: a repository having a plurality of challenges each including an interactive graphical task; a user interface for graphically presenting challenges to users; an enrollment system for assigning challenges to users and determining an authentic response for each user; an authentication system that collects an observed response from a user presented with an assigned challenge and determines a security control result based on a closeness of the observed response with the authentic response of the user.

Growing community of inventors

Bengaluru, India

Sreekrishnan Venkateswaran