公开(公告)号：WO2019103936A1

公开(公告)日：2019-05-31

申请号：PCT/US2018/061591

申请日：2018-11-16

Applicant: AMAZON TECHNOLOGIES, INC.

Inventor： SINGH, Ashish ,  SURESH, Deepikaa ,  PHILOMIN, Vasanth ,  GULABANI, Rajkumar ,  ZHUKOV, Vladimir ,  SIVASUBRAMANIAN, Swaminathan ,  ANBAZHAGAN, Vikram Sathyanarayana ,  AKARAPU, Praveen Kumar ,  STEFANI, Stefano

IPC: G10L15/30 ,  G10L15/26 ,  G10L15/187 ,  G10L15/06 ,  G10L15/183

Abstract: Techniques for automated speech recognition (ASR) are described. A user can upload an audio file to a storage location. The user then provides the ASR service with a reference to the audio file. An ASR engine analyzes the audio file, using an acoustic model to divide the audio data into words, and a language model to identify the words spoken in the audio file. The acoustic model can be trained using audio sentence data, enabling the transcription service to accurately transcribe lengthy audio data. The results are punctuated and normalized, and the resulting transcript is returned to the user.

公开(公告)号：WO2012129219A2

公开(公告)日：2012-09-27

申请号：PCT/US2012/029787

申请日：2012-03-20

Applicant: AMAZON TECHNOLOGIES, INC. ,  MARSHALL, Bradley, E. ,  SIVASUBRAMANIAN, Swaminathan ,  CERTAIN, Tate, Andrew ,  MANISCALCO, Nicholas, J.

Inventor： MARSHALL, Bradley, E. ,  SIVASUBRAMANIAN, Swaminathan ,  CERTAIN, Tate, Andrew ,  MANISCALCO, Nicholas, J.

IPC: G06F9/455

CPC classification number: G06F9/45558 ,  G06F2009/45562

Abstract: Disclosed are various embodiments for replication of machine instances in a computing environment. A clone machine instance is instantiated from a machine image associated with an original machine instance. A stored execution state of the original machine instance is applied to the clone machine instance. At least a portion of a series of stored events received by the original machine instance is applied to the clone machine instance.

Abstract translation: 公开了用于在计算环境中复制机器实例的各种实施例。 克隆机器实例从与原始机器实例相关联的机器映像实例化。 将原始机器实例的存储执行状态应用于克隆机器实例。 由原始机器实例接收的一系列存储事件的至少一部分被应用于克隆机器实例。

公开(公告)号：WO2010002603A1

公开(公告)日：2010-01-07

申请号：PCT/US2009/047832

申请日：2009-06-18

Applicant: AMAZON TECHNOLOGIES, INC. ,  SIVASUBRAMANIAN, Swaminathan ,  RICHARDSON, David, R. ,  SCOFIELD, Christopher, L. ,  MARSHALL, Bradley, E.

Inventor： SIVASUBRAMANIAN, Swaminathan ,  RICHARDSON, David, R. ,  SCOFIELD, Christopher, L. ,  MARSHALL, Bradley, E.

IPC: G06F15/16

CPC classification number: H04L61/1511 ,  G06F8/65 ,  G06F9/445 ,  G06F15/16 ,  H04L29/12066 ,  H04L29/12132 ,  H04L61/1552 ,  H04L67/02 ,  H04L67/10 ,  H04L67/2842

Abstract: A system, method and computer-readable medium for request routing. A DNS server at a content delivery network service provider obtains a DNS query corresponding to a resource requested from a client computing device and associated with a first resource identifier. The first resource identifier includes a first portion with DNS information and a second portion with path information. The DNS server selects a network computing component for processing the requested resource based on the DNS portion of the resource identifier and transmits information identifying the selected network computing component to the client computing device.

Abstract translation: 用于请求路由的系统，方法和计算机可读介质。 内容传送网络服务提供商的DNS服务器获得与从客户计算设备请求的资源相对应的与第一资源标识符相关联的资源的DNS查询。 第一资源标识符包括具有DNS信息的第一部分和具有路径信息的第二部分。 DNS服务器基于资源标识符的DNS部分选择网络计算组件来处理所请求的资源，并将识别所选择的网络计算组件的信息发送到客户端计算设备。

公开(公告)号：WO2019103978A1

公开(公告)日：2019-05-31

申请号：PCT/US2018/061869

申请日：2018-11-19

Applicant: AMAZON TECHNOLOGIES, INC.

Inventor： SWAN, Charles, Drummond ,  LIBERTY, Edo ,  LOEPPKY, Steven, Andrew ,  STEFANI, Stefano ,  SMOLA, Alexander, Johannes ,  SIVASUBRAMANIAN, Swaminathan ,  WILEY, Craig ,  BICE, Richard, Shawn ,  FAULHABER, JR., Thomas, Albert ,  GOODHART, Taylor

IPC: G06F9/50

CPC classification number: G06F9/5072 ,  G06F9/5077 ,  G06N20/00

Abstract: 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.

公开(公告)号：WO2019103931A1

公开(公告)日：2019-05-31

申请号：PCT/US2018/061522

申请日：2018-11-16

Applicant: AMAZON TECHNOLOGIES, INC.

Inventor： FAULHABER, JR., Thomas Albert ,  LIBERTY, Edo ,  STEFANI, Stefano ,  KARNIN, Zohar ,  WILEY, Craig ,  LOEPPKY, Steven Andrew ,  SIVASUBRAMANIAN, Swaminathan ,  SMOLA, Alexander Johannes ,  GOODHART, Taylor

IPC: G06N3/04

CPC classification number: G06N20/00 ,  G06N3/0454 ,  G06N3/08 ,  G06N5/04

Abstract: Techniques for dynamic accuracy-based experimentation and deployment of machine learning (ML) models are described. Inference traffic flowing to ML models and the accuracy of the models is analyzed and used to ensure that better performing models are executed more often via model selection. A predictive component can evaluate which model is more likely to be accurate for certain input data elements. Ensemble techniques can combine inference results of multiple ML models to aim to achieve a better overall result than any individual model could on its own.

公开(公告)号：WO2011053594A1

公开(公告)日：2011-05-05

申请号：PCT/US2010/054139

申请日：2010-10-26

Applicant: AMAZON TECHNOLOGIES, INC. ,  MCALISTER, Grant Alexander MacDonald ,  SIVASUBRAMANIAN, Swaminathan

Inventor： MCALISTER, Grant Alexander MacDonald ,  SIVASUBRAMANIAN, Swaminathan

IPC: G06F11/00

CPC classification number: G06F11/1469 ,  G06F11/1443 ,  G06F11/1451 ,  G06F11/1464 ,  G06F11/2025 ,  G06F11/2028 ,  G06F11/2041 ,  G06F11/2048 ,  G06F11/2056 ,  G06F11/2064 ,  G06F11/2069 ,  G06F11/2076 ,  G06F11/2082 ,  G06F17/30174 ,  G06F17/30581

Abstract: 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.

Abstract translation: 数据库环境中的复制实例提供自动故障转移和恢复。 监视组件可以定期与实例的主副本和副副本进行通信，每个实例可以驻留在单独的数据区域或地理位置以提供一定程度的可靠性和可用性。 在主实例上运行的数据库可以将信息在块级别同步复制到辅助副本，以便主副本和辅助副本同步。 如果监视组件无法与其中一个副本进行通信，则监视组件可以尝试确定这些副本是否可以相互通信，以及副本是否具有相同的数据生成版本。 根据状态信息，监视组件可以自动执行恢复操作，例如故障切换到辅助副本或执行辅助副本恢复。

公开(公告)号：WO2010114923A1

公开(公告)日：2010-10-07

申请号：PCT/US2010/029476

申请日：2010-03-31

Applicant: AMAZON TECHNOLOGIES, INC. ,  SIVASUBRAMANIAN, Swaminathan ,  MCALISTER, Grant, A.m. ,  FRANKLIN, Paul, David ,  SHETH, Rajesh, Sudhaker ,  HORSLEY, James

Inventor： SIVASUBRAMANIAN, Swaminathan ,  MCALISTER, Grant, A.m. ,  FRANKLIN, Paul, David ,  SHETH, Rajesh, Sudhaker ,  HORSLEY, James

IPC: G06F7/00

CPC classification number: G06F12/02 ,  G06F11/3006 ,  G06F11/3055 ,  G06F11/3409 ,  G06F17/30306 ,  G06Q10/06

Abstract: 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.

Abstract translation: 使用单独的控制环境来管理诸如数据存储和实例的创建，配置和管理等数据环境的各个方面。 用户可以调用控制环境的面向外部的接口，分析呼叫以确定要在数据环境中执行的动作。 控制平面的监视组件还可以周期性地与数据环境进行通信，以确定要执行的任何必要的动作，例如从数据环境中的故障​​或事件中恢复。 可以实例化工作流，其中包括执行操作所需的任务。 对于每个任务，可以将状态信息传递到可操作以执行任务的数据环境中的组件，直到完成操作的所有任务。 数据环境中的数据可以直接使用数据环境的面向外部的接口进行访问，而无需访问控制平面。

公开(公告)号：WO2020005806A1

公开(公告)日：2020-01-02

申请号：PCT/US2019/038670

申请日：2019-06-24

Applicant: AMAZON TECHNOLOGIES, INC.

Inventor： SENGUPTA, Sudipta ,  PERUMALLA, Poorna Chand Srinivas ,  DIVAKARUNI, Dominic Rajeev ,  BSHARA, Nafea ,  DIRAC, Leo Parker ,  SAHA, Bratin ,  WOOD, Matthew James ,  OLGIATI, Andrea ,  SIVASUBRAMANIAN, Swaminathan

IPC: G06F9/50

Abstract: Implementations detailed herein include description of a computer-implemented method. In an implementation, the method at least includes receiving an application instance configuration, an application of the application instance to utilize a portion of an attached accelerator during execution of a machine learning model and the application instance configuration including: an indication of the central processing unit (CPU) capability to be used, an arithmetic precision of the machine learning model to be used, an indication of the accelerator capability to be used, a storage location of the application, and an indication of an amount of random access memory to use.

公开(公告)号：WO2018231459A1

公开(公告)日：2018-12-20

申请号：PCT/US2018/034061

申请日：2018-05-23

Applicant: AMAZON TECHNOLOGIES, INC.

Inventor： ANBAZHAGAN, Vikram Sathyanarayana ,  SIVASUBRAMANIAN, Swaminathan ,  STEFANI, Stefano ,  ZHUKOV, Vladimir

IPC: G06F17/28 ,  G06F9/451 ,  G06F9/445

Abstract: Methods, systems, and computer-readable media for exporting dialog-driven applications to digital communication platforms are disclosed. A launch condition is received from a user. The launch condition is caused to be registered with one or more digital communication platforms. Detection of the launch condition is to cause a natural language input to be routed from at least one of the digital communication platforms to an application management service.

公开(公告)号：WO2011053595A1

公开(公告)日：2011-05-05

申请号：PCT/US2010/054141

申请日：2010-10-26

Applicant: AMAZON TECHNOLOGIES, INC. ,  MCALISTER, Grant, Alexander, MacDonald ,  SIVASUBRAMANIAN, Swaminathan ,  HUNTER, Barry, B., Jr. ,  BRAZIL, Silas, M.

Inventor： MCALISTER, Grant, Alexander, MacDonald ,  SIVASUBRAMANIAN, Swaminathan ,  HUNTER, Barry, B., Jr. ,  BRAZIL, Silas, M.

IPC: G06F7/00 ,  G06F17/00

CPC classification number: H04L43/10 ,  G06F11/2028 ,  G06F11/203 ,  G06F11/2035 ,  G06F11/2069 ,  G06F11/2076 ,  G06F11/2082 ,  G06F11/3006 ,  G06F11/3055 ,  G06F2201/80 ,  G06F2201/86

Abstract: Replicated instances in a database environment provide for automatic failover and recovery. A monitoring component can obtain a lease enabling the component to periodically communicate with, and monitor, one or more data instances in the data environment, where the data instance can be a replicated instance including a primary and a secondary replica. For a large number of instances, the data environment can be partitioned such that each monitoring component can be assigned a partition of the workload. In the event of a failure of a monitoring component, the instances can be repartitioned and the remaining monitoring components can be assigned to the new partitions to substantially evenly distribute the workload.

Abstract translation: 数据库环境中的复制实例提供自动故障转移和恢复。 监视组件可以获得租赁，使组件能够周期性地与数据环境中的一个或多个数据实例进行通信并监视数据实例，其中数据实例可以是包含主副本和辅助副本的复制实例。 对于大量实例，可以对数据环境进行分区，以便每个监视组件都可以分配工作负载分区。 如果监视组件发生故障，则可以重新分区实例，并将其余监视组件分配给新分区，以基本上均匀分布工作负载。