Amazon Technologies, Inc.

Gourav Roy

Srujana Merugu

Gowda Dayananda Anjaneyapura Range

Aswin Natarajan

Amit Chandak

Bharath Venkatesh

Prateek Gupta

Sathish Kumar Palanisamy

Jagannathan Srinivasan

Ritesh Singh

Marc John Brooker

Anthony Nicholas Liguori

Gaurav D Ghare

Roger Shane Barga

Vineet Khare

William J Vass

Vineet Shashikant Chaoji

Fusheng Yuan

Saurabh Sohoney

Kevin Robert McGehee

Marc Bowes

Bin Zhu

Juan Carlos Gomez

Dmitry Berkovich

Qu Chen

Iftach Ragoler

Abhishek Kumar

Chetan Banthiya

Andrew Dennis Willingham

Ahmed Hesham Salem

Jan Raak

Itay Maoz

Allen Robert Samuels

James Alexander Morle

Paul Andrew Lafranchise

Swarnaprakash Udayakumar

Brandon Kyle Pelfrey

Sivaramakrishnan Kaveri

Jason Chen Hooi Chionh

Compact hosting of database query processors is performed using virtualization snapshots. When a request to access a database is received via a previously established connection from a client for an idle query processor that is not currently executing, a virtualization snapshot may be identified that is associated with the idle query processor. The virtualization snapshot of the query processor may be used to restore the query processor into virtual execution on a host system for the database using an origin snapshot that shares clean pages across multiple query processors on the host system. The request may be performed by the restored query processor in virtual execution on the host system.

Compact hosting of database query processors using virtualization snapshots

A database management system maintains a collection of data using a log-based storage structure. In response to a request to store data items (key-value pairs), the database management system stores mapping information to a data item in an in-memory data structure while the data items are stored on the log-based storage structure. The hash of the key includes information to identify an index entry that comprises information that maps to the location of the data item stored on the log-based storage structure.

Database with in-memory data stores

A robotic device management service obtains, from a client device operating in a first network, a request to obtain data from a robotic device operating in a second network. In response to the request, the robotic device management service issues a token to the client device that can be provided in future queries to obtain the data. The robotic device management service provides parameters of the request to the robotic device to cause the robotic device to generate and provide the data to the robotic device management service. In response to another request to obtain the data, where the other request includes the token, the robotic device management service queries a database to determine whether the data is available from a storage location of the service. If the data is available, the service provides the data to the client device to fulfill the other request.

Robotics application development and monitoring over distributed networks

A robotic device management service obtains, from a customer, a first set of parameters of a robotic device and a second set of parameters for a simulation environment for testing a robotic device application installable on the robotic device. The set of parameters are used to indicate a storage location of the application and a selection of a simulation environment for testing the application. In response to the request, the robotic device management service selects a set of resources on which to execute the simulation in the simulation environment. The robotic device management service obtains the robotic device application from the storage location and loads the application on to the set of resources to execute the simulation.

Robotics application development architecture

A robotic device management service obtains a request from a client device to cause a fleet of robotic devices operating at a remote facility to perform a set of tasks. In response to obtaining the request, the robotic device management service stores the request in a queue associated with the fleet of robotic devices. A distribution device at the facility queries the queue to obtain the request and, in response to obtaining the request, schedules performance of the set of tasks by the fleet of robotic devices.

Request distribution and oversight for robotic devices

Features are provided for the analysis of collections of data and automatic grouping of data having certain similarities. A collection of data regarding user interactions with item-specific content can be analyzed. The analysis can be used to identify groups of items that are of interest to groups of similar users and/or to identify groups of users with demonstrated interests in groups of similar items. Data may be analyzed in a 'bottom-up' manner in which correlations within the data are discovered in an iterative manner, or in a 'top-down' manner in which desired top-level groups are specified at the beginning of the process. A bottom-up process may also be distributed among multiple devices or processors to more efficiently discover groups when using large collections of data.

Distributed grouping of large-scale data sets

A number of attributes of different attribute types, to be used to assign observation records of a data set to clusters, are identified. Attribute-type-specific distance metrics for the attributes, which can be combined to obtain a normalized aggregated distance of an observation record from a cluster representative, are selected. One or more iterations of a selected clustering methodology are implemented on the data set using resources of a machine learning service until targeted termination criteria are met. A given iteration includes assigning the observations to clusters of a current version of a clustering model based on the aggregated distances from the cluster representatives of the current version, and updating the cluster representatives to generate a new version of the clustering model.

Scalable clustering for mixed machine learning data

An approximate data structure to represent clusters of observation records of a data set is identified. A hierarchical representation of a plurality of clusters, including the targeted number of clusters among which the observation records are to be distributed, is generated. Each node of the hierarchy comprises an instance of the approximate data structure. Until a set of termination criteria are met, iterations of a selected clustering methodology are run. In a given iteration, distances of observation records from the cluster representatives of a current version of the model are computed using the hierarchical representation, and a new version of the model with modified cluster representatives is generated.

Clustering sparse high dimensional data using sketches

Growing community of inventors

Seattle, WA, United States of America

Gourav Roy