Amazon Technologies, Inc.

Swaminathan Sivasubramanian

Bradley E Marshall

David R Richardson

Tate Andrew Certain

Grant Alexander Macdonald McAlister

Christopher Laurence Scofield

Eric Jason Brandwine

Stefano Stefani

Timothy Andrew Rath

Tal Saraf

Nicholas J Maniscalco

Elmore Eugene Pope

Rajesh Sudhakar Sheth

Rande Anthony Blackman

Raymond S Bradford

Srinivasan Sundar Raghavan

Mukul Vijay Karnik

Santosh Kalki

Dan Mihai Dumitriu

Grant A M McAlister

Amol Devgan

Jakub Kulesza

Vikram Sathyanarayana Anbazhagan

Luis Felipe Cabrera

James Robert Hamilton

Wei Xiao

John David Cormie

Paul David Franklin

Khawaja Salman Shams

Chiranjeeb Buragohain

Prashant Pandey

James Horsley

Nafea Bshara

Sudipta Sengupta

Leo Parker Dirac

Andrea Olgiati

Peter Sven Vosshall

Omer Ahmed Zaki

Poorna Chand Srinivas Perumalla

Matthew James Wood

Barry Bailey Hunter, Jr

Vladimir Zhukov

Deniz Hastorun

Alex Pilchin

Ivan D Rosero

Dominic Rajeev Divakaruni

Bratin Saha

Matthew Scott Garman

David Craig Yanacek

Johanna S Olson

Sachin Jain

Adam Douglas Morley

Thomas Albert Faulhaber, Jr

Alexander Johannes Smola

Milovan Milovanovic

Giuseppe Decandia

Craig Wiley

Avinash Lakshman

Vasanth Philomin

Johnny Ying Wu

Marvin Michael Theimer

Alan Michael Judge

Don Johnson

Brad Eugene Marshall

Imran S Patel

Parikshit Shivajirao Pol

Ashish Singh

Silas M Brazil

Vivek Rai

Edo Liberty

Rama Krishna Sandeep Pokkunuri

Zohar Karnin

Gowda Dayananda Anjaneyapura Range

Taylor Goodhart

Atul Deo

Chiranjeeb Buragahain

Steven Andrew Loeppky

Jessie Young

Anuroop Arora

Allan Henry Vermeulen

Pradeep Vincent

Yi Zhang

Rahul Sharma

Gokul Soundararajan

Maximiliano Maccanti

Jai Vasanth

Benjamin Aldouby Schwartz

David Robert Thomas

Ganesh Kumar Gella

Yahor Pushkin

Yaser Al-Onaizan

Richard Shawn Bice

Roger Scott Jenke

Harshal Pimpalkhute

Charles Drummond Swan

Meghana Puvvadi

Grant Alexander MacDonald McAlister

Benjamin Ws Redman

Jalaja C Kurubarahalli

Aanchal Gupta

A system and method for management and processing of resource requests is provided. A content delivery network service provider receives a DNS query from a client computing device. The DNS query corresponds to a requested resource from the client computing device. The content delivery network service provider associates the client computing device with a cluster of other client computing devices. Based on routing information for the cluster, the content delivery network service provider routes the DNS query. The process can further include monitoring performance data associated with the delivery of the requested resource and updating the routing information for the cluster based on the performance data for use in processing subsequent requests from client computing devices in the cluster.

Request routing based on class

A system that implements a scaleable data storage service may maintain tables in a data store on behalf of storage service clients. The service may maintain data in partitions stored on respective computing nodes in the system. The service may support multiple throughput models, including a committed throughput model and a best effort throughput model. A service request to create a table may specify that requests directed to the table should be serviced under a committed throughput model and may specify the committed throughput level in terms of logical service request units. The service may reserve low-latency storage and other resources sufficient to meet the specified committed throughput level. A client/user may request a modification to the committed throughput level in anticipation of workload changes, such as an increase or decrease in traffic or data volume. In response, the system may increase or decrease the resources reserved for the table.

System and method for providing a committed throughput level in a data store

A system that implements a scalable data storage service may maintain tables in a non-relational data store on behalf of clients. The system may provide a Web services interface through which service requests are received, and an API usable to request that a table be created, deleted, or described; that an item be stored, retrieved, deleted, or its attributes modified; or that a table be queried (or scanned) with filtered items and/or their attributes returned. An asynchronous workflow may be invoked to create or delete a table. Items stored in tables may be partitioned and indexed using a simple or composite primary key. The system may not impose pre-defined limits on table size, and may employ a flexible schema. The service may provide a best-effort or committed throughput model. The system may automatically scale and/or re-partition tables in response to detecting workload changes, node failures, or other conditions or anomalies.

System and method for implementing a scalable data storage service

Replicated instances in a database environment provide for automatic failover and recovery. A monitoring component can periodically communicate with a primary and a secondary replica for an instance, with each capable of residing in a separate data zone or geographic location to provide a level of reliability and availability. A database running on the primary instance can have information synchronously replicated to the secondary replica at a block level, such that the primary and secondary replicas are in sync. In the event that the monitoring component is not able to communicate with one of the replicas, the monitoring component can attempt to determine whether those replicas can communicate with each other, as well as whether the replicas have the same data generation version. Depending on the state information, the monitoring component can automatically perform a recovery operation, such as to failover to the secondary replica or perform secondary replica recovery.

Failover and recovery for replicated data instances

Techniques for in-flight scaling of machine learning training jobs are described. A request to execute a machine learning (ML) training job is received within a provider network, and the ML training job is executed using a first one or more compute instances. Upon a determination that a performance characteristic of the ML training job satisfies a scaling condition, a second one or more compute instances are added to the ML training job while the first one or more compute instances continue to execute portions of the ML training job.

In-flight scaling of machine learning training jobs

Aspects of a data environment, such as the creation, provisioning, and management of data stores and instances, are managed using a separate control environment. A user can call into an externally-facing interface of the control environment, the call being analyzed to determine actions to be performed in the data environment. A monitoring component of the control plane also can periodically communicate with the data environment to determine any necessary actions to be performed, such as to recover from faults or events in the data environment. A workflow can be instantiated that includes tasks necessary to perform the action. For each task, state information can be passed to a component in the data environment operable to perform the task, until all tasks for an action are completed. Data in the data environment can be accessed directly using an externally-facing interface of the data environment, without accessing the control plane.

Control service for data management

A determination is made as to whether a value of a first parameter of a first application is to be obtained using a natural language interaction. Based on received input, a first service of a plurality of services is identified. The first service is to be used to perform a first task associated with the first parameter. Portions of the first application to determine the value of the first parameter and to invoke the first service are generated.

Back-end task fulfillment for dialog-driven applications

A system receives one or more transcripts of communications between entities. The system identifies a requested action in the communications based at least in part on a mapping between the requested action and an application programming interface. The system identifies one or more statements eliciting information, based on parameters to the application programming interface. The system generates a definition of an artificial agent based, at least in part, on the requested action and the one more statements eliciting information.

Artificial agent generator

With the advent of virtualization technologies, networks and routing for those networks can now be simulated using commodity hardware. For example, virtualization technologies can be adapted to allow a single physical computing machine to be shared among multiple virtual networks by providing one or more virtual machines simulated in software by the single physical computing machine, with each virtual machine acting as a distinct logical computing system. In addition, as routing can be accomplished through software, additional network setup flexibility can be provided to the virtual network in comparison with hardware-based routing. In some implementations, virtual network setup can be abstracted through the use of resource placement templates, allowing users to create virtual networks compliant with a customer's networking policies without necessarily having knowledge of what those policies are.

Resource placement templates for virtual networks

Techniques for using scoring algorithms utilizing containers for flexible machine learning inference are described. In some embodiments, a request to host a machine learning (ML) model within a service provider network on behalf of a user is received, the request identifying an endpoint to perform scoring using the ML model. An endpoint is initialized as a container running on a virtual machine based on a container image and used to score data and return a result of said scoring to a user device.

Packaging and deploying algorithms for flexible machine learning

Implementations detailed herein include description of a computer-implemented method. In an implementation, the method at least includes attaching a first set of one or more accelerator slots of an accelerator appliance to an application instance of a multi-tenant provider network according to an application instance configuration, the application instance configuration to define per accelerator slot capabilities to be used by an application of the application instance, wherein the multi-tenant provider network comprises a plurality of computing devices configured to implement a plurality of virtual compute instances, and wherein the first set of one or more accelerator slots is implemented using physical accelerator resources accessible to the application instance; while performing inference using the loaded machine learning model of the application using the first set of one or more accelerator slots on the attached accelerator appliance, managing resources of the accelerator appliance using an accelerator appliance manager of the accelerator appliance.

Fault-tolerant accelerator based inference service

A replicated database can be provisioned that provides primary and secondary replicas that can be provisioned in different data zones or geographical locations. The database can be installed on the primary replica, and both the primary and secondary replica can have installed a block level replication mechanism that allows any I/O operation to be replicated by between the primary and secondary replicas. Any failure or outage of the primary replica can be addressed by performing a failover operation to the secondary replica. A DNS name or other such approach can be used such that the name can be aliased to the secondary replica during a failover, such that there is no action needed on the part of the customer to utilize the 'new' primary replica. The creation of the database and provisioning of the replicated instance can be initiated using a Web service call to a control environment. A replicated database can also be scaled according to storage or computing capacity with no disruption of service using a Web service call to the control environment.

Provisioning and managing replicated data instances

Systems and methods to implement customer contact service with real-time supervisor assistance. A supervisor may oversee multiple agents in a customer contact service. A service of a computing resource service provider may monitor a plurality of audio connections at a service of a computing resource service provider, generate transcripts for the plurality of audio data, analyze the transcripts using a set of natural language processing (NLP) techniques to generate metadata, tag the transcripts with categories based at least in prat on the metadata, generate information for at least a portion of the plurality of connections based on the transcripts, metadata, and categories, and provide the information to a supervisor of the agents.

Customer contact service with real-time supervisor assistance

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

Growing community of inventors

Seattle, WA, United States of America

Swaminathan Sivasubramanian