Abstract: The present disclosure provides a system and method rendering enhanced multi-category query responses via orchestrated execution of trained artificial intelligent (AI) agent instances. An execution control module can receive user queries via a user interface and direct a query response module to determine semantic categories associated with the queries, generate corresponding subtasks, and select trained agent instances to execute the subtasks. Subtask responses can be validated and combined to generate a structured responses to the queries. The structured responses can be validated and presented to the users via the user interface.

Abstract: Systems, methods, and computer-readable storage media for reinforcing and modifying the search engine behavior using aspects of federated learning. A system can receive, from a plurality of first users at a computer system, a query for a product, and generate a vector representation of the query. The system can then execute a search of an embedding space for the vector representation, resulting in a first ranked list of at least one product, wherein the embedding space comprises vectorized product representations for a plurality of products, and wherein the first ranked list is ranked in order of similarity of the vectorized product representations to the vector representation of the query. As feedback is received from multiple users, and the query is received again, the system can generate a second ranked list which is the first ranked list modified by the feedback of the plurality of first users.

Abstract: Systems, methods, and computer-readable storage media for graph model generation, and more specifically to generating graph models using supply chain data and operations. A system can receive, from a plurality of sources, sensor data, each piece of the sensor data including information associated with an exchange. The system can then parse, via at least one processor, the sensor data to identify components of each piece of the sensor data, resulting in parsed sensor data. The system resolves, via the processor, missing data within the parsed sensor data, resulting in parsed, resolved sensor data. The system can then map, via the at least one processor, the parsed, resolved sensor data to a graph data structure, the graph data structure having nodes and edges, and store the graph data structure in a graph database.

Abstract: Systems, methods, and computer-readable storage media for reinforcing and modifying the search engine behavior using aspects of federated learning. A system can receive from a user at a first time, a query for a product, then generate a vector representation of the query, and execute a search of an embedding space for the vector representation, resulting in at least one similarity score for at least one product. The system can present to the user search results based on the at least one similarity score, and receive from the user a selected product from within the search results. The system can update, based on the selected product, the at least one similarity score for the selected product, resulting in at least one updated similarity score, where when the query is received again the search of the embedding space for the vector representation produces the at least one updated similarity score.

Abstract: Systems, methods, and computer-readable storage media for reinforcing and modifying the search engine behavior using aspects of federated learning. A system can receive from a user at a first time, a query for a product, then generate a vector representation of the query, and execute a search of an embedding space for the vector representation, resulting in at least one similarity score for at least one product. The system can present to the user search results based on the at least one similarity score, and receive from the user a selected product from within the search results. The system can update, based on the selected product, the at least one similarity score for the selected product, resulting in at least one updated similarity score, where when the query is received again the search of the embedding space for the vector representation produces the at least one updated similarity score.

Abstract: Systems, methods, and computer-readable storage media for reinforcing and modifying the search engine behavior using aspects of federated learning. A system can receive, from a plurality of first users at a computer system, a query for a product, and generate a vector representation of the query. The system can then execute a search of an embedding space for the vector representation, resulting in a first ranked list of at least one product, wherein the embedding space comprises vectorized product representations for a plurality of products, and wherein the first ranked list is ranked in order of similarity of the vectorized product representations to the vector representation of the query. As feedback is received from multiple users, and the query is received again, the system can generate a second ranked list which is the first ranked list modified by the feedback of the plurality of first users.

Abstract: Aspects of the subject technology relate to systems, methods, and computer-readable media for predicting energy states of devices operating in an environment. An operational constraint and a device setting for a device operating in an environment can be identified. An environmental state of the environment can be predicted based on the device operating at the device setting and under the operational constraint through application of an environmental state model that maps varying operational constraints and varying device settings to varying environmental states in the environment. An energy state for operating the device in the environment at the device setting and under the operational constraint can be predicted based on the predicted environmental state through application of an energy consumption model that maps the varying environmental states to varying energy states.

Abstract: A computer method for identifying product in a distributor's inventory system that fulfills a product request made via a natural language query. the natural language query is received as a product request including multiple words in sequential order. The words are vectorized into word-vectors that are concatenated and used to generate a query embedding. The query embedding is processed utilizing a trained product category classifier that predicts which product category the requested product belongs. Forward and backward sequence vectors are generated from the sequentially ordered words of the query that are concatenated and processed using a trained model specific to the predicted product category. The sequence vectors represent positional relationships between the words of the natural language query. Thereafter, the system identifies product attribute(s) embodied in the natural language query that each correspond to a predetermined key-characteristic of the category.

Abstract: System and methods for training neural network models for real-time flow simulations are provided. Input data is acquired. The input data includes values for a plurality of input parameters associated with a multiphase fluid flow. The multiphase fluid flow is simulated using a complex fluid dynamics (CFD) model, based on the acquired input data. The CFD model represents a three-dimensional (3D) domain for the simulation. An area of interest is selected within the 3D domain represented by the CFD model. A two-dimensional (2D) mesh of the selected area of interest is generated. The 2D mesh represents results of the simulation for the selected area of interest. A neural network is then trained based on the simulation results represented by the generated 2D mesh.

Abstract: Aspects and features of this disclosure relate to projecting physical drilling parameters to control a drilling operation. A computing system applies Bayesian optimization to a model incorporating the input data using varying values for an adverse drilling factor to produce a target function. The computing system determines a minimum value for the target function. The computing system provides a projected value for the physical drilling parameters based on the minimum value. The computing system generates an alert responsive to determining that the projected value for the physical drilling parameters exceeds a prescribed limit.

Abstract: A computer implemented method for identifying substitute products for a target product via a distributor's product management computer system, the method includes: receiving, at a distributor's product management computer system, data representing a target product of interest to a user of the system, wherein the target product is described by a plurality of variable-value attributes associated therewith; processing the received data through a machine-learning-produced algorithm and thereby generating data representing a first set of substitute product candidates for the target product and wherein each substitute product candidate has a corresponding match-score that represents a degree of determined similarity between that substitute product candidate and the target product; transforming the generated data that represents the first set of substitute product candidates into data representing a refined set of substitute product candidates by processing the generated data utilizing a constraint-based algorithm;

Abstract: A product management computer system that can include one or more processors, and a computer-readable medium comprising instructions stored therein, which when executed by the processors, can cause the processors to: receive data representing a target product of interest to a user of the system, wherein the target product is described by a plurality of variable-value attributes associated; process the received data through a machine-learning-produced algorithm and generating data representing a first set of substitute product candidates for the target product and wherein each substitute product candidate has a corresponding match-score that represents a degree of determined similarity between that substitute product candidate and the target product; transform the generated data that represents the first set of substitute product candidates into data representing a refined set of substitute product candidates by processing the generated data utilizing a constraint-based algorithm; and output data representing

Abstract: A product management computer system that can include one or more processors, and a computer-readable medium comprising instructions stored therein, which when executed by the processors, can cause the processors to: receive data representing a target product of interest to a user of the system, wherein the target product is described by a plurality of variable-value attributes associated; process the received data through a machine-learning-produced algorithm and generating data representing a first set of substitute product candidates for the target product and wherein each substitute product candidate has a corresponding match-score that represents a degree of determined similarity between that substitute product candidate and the target product; transform the generated data that represents the first set of substitute product candidates into data representing a refined set of substitute product candidates by processing the generated data utilizing a constraint-based algorithm; and output data representing

Abstract: A computer method for identifying product in a distributor's inventory system that fulfills a product request made via a natural language query. the natural language query is received as a product request including multiple words in sequential order. The words are vectorized into word-vectors that are concatenated and used to generate a query embedding. The query embedding is processed utilizing a trained product category classifier that predicts which product category the requested product belongs. Forward and backward sequence vectors are generated from the sequentially ordered words of the query that are concatenated and processed using a trained model specific to the predicted product category. The sequence vectors represent positional relationships between the words of the natural language query. Thereafter, the system identifies product attribute(s) embodied in the natural language query that each correspond to a predetermined key-characteristic of the category.

Abstract: Systems, methods, and computer-readable media are described for intelligent, real-time monitoring and managing of changes in oilfield equilibrium to optimize production of desired hydrocarbons and economic viability of the field. In some examples, a method can involve generating, based on a topology of a field of wells, a respective graph for the wells, each respective graph including computing devices coupled with one or more sensors and/or actuators. The method can involve collecting, via the computing devices, respective parameters associated with one or more computing devices, sensors, actuators, and/or models, and identifying a measured state associated with the computing devices, sensors, actuators, and/or models.

Abstract: Aspects and features of this disclosure relate to projecting physical drilling parameters to control a drilling operation. A computing system applies Bayesian optimization to a model incorporating the input data using varying values for an adverse drilling factor to produce a target function. The computing system determines a minimum value for the target function. The computing system provides a projected value for the physical drilling parameters based on the minimum value. The computing system generates an alert responsive to determining that the projected value for the physical drilling parameters exceeds a prescribed limit.

Abstract: Systems, methods, and computer-readable media for a well construction activity graph builder. An example method can include obtaining a stream of events associated with a wellbore; obtaining mapping metadata identifying data points to be included in a graph data model from a store of data associated with the wellbore; generating the graph data model based on the stream of events, the mapping metadata, and the data points identified in the mapping metadata, the graph data model including nodes representing logical entities associated with the data points, the nodes having interconnections based on data relationships defined in the mapping metadata, each logical entity corresponding to a set of data points from the data points; and generating a view of the graph data model, the view depicting at least some of the nodes and interconnections in the graph data model.

Abstract: An automated offset well analytics engine generates offset well rankings for a prospect well. The engine aggregates data for offset wells and the prospect wells is across multiple disparate data sources corresponding to a user-specified scope. The engine generates features comparing the offset wells to the prospect well are using a combination of machine-learning based models and risk analysis. Offset wells are ranked by feature and further ranked across features using a weighted feature ranking map. Feature weights are iteratively trained using a reinforcement learning model in a feedback loops with a well expert. A prospect well casing schema and bottom hole assembly is designed using automatically generated offset well rankings.

Abstract: Systems and methods for auto-detection and classification of rig activities from trend analysis of sensor data are provided. Sensor data from rig equipment may be obtained during wellsite operations. The sensor data may be analyzed to identify one or more index points where a trend in the sensor data changes. The sensor data may be segmented into a first set of time segments representing macro activities performed during the well site operations, based on the one or more identified index points. Statistical analysis may be performed on the sensor data within each first time segment to identify points where statistical properties of the sensor data change. Each first time segment may he segmented into a second set of time segments representing micro activities performed during the wellsite operations, based on the identified points of change in the statistical properties.

Abstract: A drill bit subsystem can include a drill bit, a processor, and a non-transitory computer-readable medium for storing instructions and for being positioned downhole with the drill bit. The instructions of the non-transitory computer-readable medium can include a machine-teachable module and a control module that are executable by the processor. The machine-teachable module can receive depth data and rate of drill bit penetration from one or more sensors in a drilling operation, and determine an estimated lithology of a formation at which the drill bit subsystem is located. The control module can use the estimated lithology to determine an updated location of the drill bit subsystem, and control a direction of the drill bit using the updated location and a drill plan.