Amazon Technologies, Inc.

Srinivasan Sundar Raghavan

Timothy Andrew Rath

Santosh Kalki

Swaminathan Sivasubramanian

Mukul Vijay Karnik

Amol Devgan

Anurag Windlass Gupta

Hersh Nagar

Douglas Scott Cerny

Stephanie Weeks

Daniel Micol-Ponce

Vijaya Prasad Reddy Avula

Noelia Moron

Jaime Vallori

Vijay Chaudhari

Vamsi Krishna Vutukuru

Anurang Windlass Gupta

A platform for data analytics may be provided in a hosted environment on a multi-tenant system. The platform provider may also provide transactional processing services. Data obtained from processing the transactional services may be stored in an n-dimensional cube with which analytics may be performed. A dimension and hierarchy model may be identified based on correlations between hierarchy dimensions and levels in a dataset, or in schema and queries related to the dataset. Correlations may be further based on data received from a data stream. Priority for calculating a hierarchy may be based on data received from a data stream.

Automated hierarchy detection for cloud-based analytics

Devices and techniques are generally described for dynamic policy determination using machine learning. In some examples, first goal data may be received. A first machine learning model may generate a first computer-executable policy based at least in part on the first goal data. In some examples, a first search query may be received. In various examples, the first search query may be modified by the first computer-executable policy to generate a modified search query. In some examples, first feedback data related to user interaction with search results of the modified search query may be determined. In various examples, at least one parameter of the first machine learning model may be updated based at least in part on the first feedback data to generate a second machine learning model.

Machine learning system for dynamic generation of computer-implemented policies

Devices and techniques are generally described for a speech processing procurement system. In various examples, recommendations may be provided by back-end systems. The recommendations may be ranked by a machine learning model, and may be sorted and filtered for a particular output device. Various voice guidance workflows may be invoked through user interaction with a speech processing procurement system. The workflows may be used to surface pre-recorded voice memos and/or to determine recommendations based on dynamically updated procurement policies.

Speech processing procurement system

A transactional data storage engine may implement selectable storage endpoints. A selection of storage endpoints may be received at a transactional data storage engine. The selected storage endpoints may identify storage locations maintaining replicas of data for the transactional data storage engine. A storage engine configuration for the transactional data storage engine may be updated to include the storage endpoints so that access requests for the data may be sent to storage endpoints identified according to the storage engine configuration. In some embodiments, storage endpoints may identify strongly consistent or eventually consistent storage locations for performing reads of the data maintained for the transactional data storage engine.

Selectable storage endpoints for a transactional data storage engine

A probabilistic counting structure such as a hyperloglog may be formed during a table scan for each of a selected set of columns. The columns may be selected based on an initial estimate of relatedness, which may be based on data types of the respective columns. An estimated cardinality of an intersection or union of columns may be formed based on an intersection of the probabilistic data structures. A join path may be determined based on the estimated cardinality of an intersection or union of the columns.

Efficient determination of join paths via cardinality estimation

A hosted analytics system may be integrated with transactional data systems and additional data sources such real-time systems and log files. A data processing pipeline may transform data on arrival for incorporation into an n-dimensional cube. Correlation between patterns of events in transactional data may be identified. Upon arrival, new data may be transformed and incorporated into the n-dimensional cube. Similarity between the new data and a previously identified correlation may be determined and flagged.

Dynamic detection of data correlations based on realtime data

An online analytical processing system may comprise an n-dimensional cube partitioned into slices, in which each slice may represent data points at the intersections of fixed and variable dimensions. Computation of data points within a slice may be deferred. A dependency graph may be initially constructed, in which the dependency graph is utilized in a subsequent computation. Calculation of data points may be prioritized based on information indicative of a chance that the data points will be accessed.

Data interest estimation for n-dimensional cube computations

Real-time hosted system analytics

A corpus of information describing queries used to access a transactional data store may be used to identify analytical relationships that are not explicitly defined in a schema or supplied by a user. Join relationships may be identified based on field coincidence in elements of queries in the corpus. Join relationships may be indicative of dimensions and attributes of a dimension. Hierarchy levels for a dimension may be identified based on factors including data type, reference in an aggregating clause, and reference in a grouping clause.

Identifying join relationships based on transactional access patterns

An online analytical processing system may comprise an n-dimensional cube structured using slice-based partitioning in which each slice comprises one or more hierarchies of data points. A region of a hierarchy may be classified according to computational demands associated with the region. A scaling or replication mechanism may be applied to the region based on the computational demands associated with that region.

Dynamic n-dimensional cubes for hosted analytics

An analytics module may be embedded into an application developed, published, or used by an entity in addition to the owner of the data under analysis. An access token may be submitted by the analytics module to a provider of hosted services. The access token may correspond to an n-dimensional cube containing data at a level of granularity permitted to the application. The access token may incorporate additional policies controlling access to the corresponding n-dimensional cube.

Embeddable cloud analytics

A multi-tenant system for providing hosted analytic services may be dynamically configured in response to a request from a user. A request for analytic services may comprise an indication of at least one data source to be incorporated into an n-dimensional cube. A data source connector and transformation pipeline may transform data received from the data source to a format compatible with a dimension and hierarchy model of the n-dimensional cube.

Growing community of inventors

Mercer Island, WA, United States of America

Srinivasan Sundar Raghavan