Abstract: A robot operates using electrical power. The robot is able to autonomously return to a dock to recharge its batteries. The robot may perform a variety of tasks that consume electrical power, such as performing video calls, presenting audio or video content, acting as a sentry, and so forth. Usage of the robot is analyzed to predict tasks that a user is likely to request. This information is used along with information about previously scheduled tasks and system tasks to create a list of tasks to be performed and when to perform those tasks. The tasks to be performed are calculated to maximize availability of the robot to the user while minimizing charge/discharge cycles, minimizing the extent of the charge/discharge, and minimizing the use of rapid charging. As a result, the lifespan of the batteries may be increased.

Abstract: An autonomous mobile device (AMD) may be used in an environment as a communication endpoint for voice or video communications. An incoming request for communication may initiate a process in which the AMD finds a user within the environment. Information obtained from sensors onboard the AMD or in the environment may be used to determine the whereabouts of the user. If an existing communication endpoint is not available to the user or cannot support a requested communication modality, the AMD may travel to permitted areas within the environment to find the user, while avoiding areas designated as private. Once found, communication may be established with the user. If the incoming request expires, the AMD may present information indicative of the request to the user.

Abstract: An autonomous mobile device (AMD) moves within a physical space. Various wireless devices with transmitters may be present, such as internet of things (IoT) devices, smartphones, tablets, user devices, and so forth. At different physical locations in the physical space a radio receiver of the AMD receives radio signals sent by transmitters sending data. Data is stored that is indicative of the physical location where the data was acquired, an identifier of the transmitter, and a received signal strength indication (RSSI). Estimated distances, each with respect to a different physical location, are determined based on the received signal strength. A plurality of the estimated distances and associated physical locations are used to determine an estimated location of a transmitter. For example, the estimated transmitter location may be determined as a point where the estimated distances coincide.

Abstract: Techniques for implementing multiple wakeword detectors on a single device are described. A digital signal processor (DSP) of the device may implement a wakeword detection component to detect when captured speech includes a wakeword. A companion application installed on the device may implement a wakeword detection component trained using speech of a user of the device. If the DSP's wakeword detection component detects a wakeword in speech, the companion application's wakeword detection component may be used to determine whether the wakeword was spoken by the user of the device. If the companion application's wakeword detection component determines the user spoke the wakeword, audio data representing the speech may be sent to at least one server(s) for processing.

Abstract: Techniques for streaming real-time automated speech recognition (ASR) are described. A user can stream audio data to a frontend service of the ASR service. The frontend service can establish a bi-directional connection to an audio decoder host to perform ASR on the data stream. The audio decoder host may include a streaming ASR engine which can analyze chunks of the audio data stream using an acoustic model to divide the audio data into words, and a language model to identify sentences made of the words spoken in the audio file. The acoustic model can be trained using short audio sentence data (e.g., on the order of 30 seconds to a few minutes), enabling the transcription service to accurately transcribe short chunks of audio data. The results are then punctuated and normalized. The resulting transcript is then streamed back to the user over the bi-directional connection.

Abstract: Customers can utilize resources of a multi-tenant environment to provide one or more services available to various users. In order to simplify the process for these customers, the multi-tenant environment can include an infrastructure wherein a portion of the resources provide an authentication and/or authorization service that can be leveraged by the customer services. These resources can logically sit in front of the resources used to provide the customer services, such that a user request must pass through the authorization and authentication service before being directed to the customer service. Such resources can provide other functionality as well, such as load balancing and metering.

Abstract: A gift privacy module receives a request to display an order history associated with a user account. In response to the request, the gift privacy module displays the order history associated with the user account, while hiding at least a portion of order details associated with at least one entry identified as private in the order history.

Abstract: Techniques described herein include a platform for generating customized media content to a user. In some embodiments, a service provider may receive attribute data related to an environment in which a media file is to be presented. In some embodiments, the service provider may receive a user configuration setting indicating one or more preferences for media presentation. The service provider may store media content as separate components, with each component being separately adjustable. Upon receiving a request for particular media content, the service provider may generate a media file from the stored media content in accordance with the user configuration settings and/or environment attribute data.

Abstract: Based on network route tracing data from a set of monitored computing nodes, pairs of network analysis agents that are allocated to monitored computing nodes that are linked by at least a target number of non-redundant network paths are identified. The identified pairs of agents are de-allocated from the set of monitored computing nodes. New pairs of agents are allocated to the set of monitored computing nodes to replace the de-allocated pairs of agents.

Abstract: Provided are systems and methods for reliable, out-of-order transmission of packets. In some implementations, provided is an apparatus configured to communicate with a network and a host device. The apparatus may receive messages from the host device at a send queue, where each message includes destination information. The apparatus may further determine, using the destination information and an identify of the send queue, a transport context associated with a destination on the network. The apparatus may further, for each message and using the transport context, generate a packet including the message and transmit the packet over the network. The apparatus may further monitor status for each transmitted packet.

Abstract: An attachment received from a webclient for a message in composition can be saved by an HTTP server in a storage system. Location information for the attachment can be stored in a database in a record associated with the webclient. If the HTTP server is unexpectedly unavailable, a backup HTTP server can locate the attachment using the record in the database and provide the attachment to a messaging server when the user is ready to send the message.

Abstract: A service allowing for obfuscation of identifiers such that the obfuscated identifier values are within a set of valid identifier values. The service allows for resources to be associated with an obfuscated identifier, and for clients to request information regarding obfuscated identifiers and resources associated to obfuscated identifiers.

Abstract: Techniques described and suggested herein include systems and methods for precomputing regeneration information for data archives (“archives”) that have been processed and stored using redundancy coding techniques. For example, regeneration information, such as redundancy code-related matrices (such as inverted matrices based on, e.g., a generator matrix for the selected redundancy code) corresponding to subsets of the shards, is computed for each subset and, in some embodiments, stored for use in the event that one or more shards becomes unavailable, e.g., so as to more efficiently and/or quickly regenerate a replacement shard.

Abstract: Architectures and techniques are described to provide a number of options to exchange information related to deals via a plurality of channels. Each of the communication channels may be utilized to exchange communications about different aspects of acquiring and redeeming deals. The channel options may be related to categories of computing devices, operating systems executed by computing devices, one or more sites, various forms of communication, client device applications, etc. A service provider that offers deals on behalf of merchants may determine one or more options for each communication channel with respect to merchants offering deals and with respect individuals that may participate in deals offered by the service provider. After determining the channel options for a deal offered by a particular merchant and for individuals designated to receive information about the deal, communications with respect to the deal may be exchanged over the channels via certain channel options.

Abstract: Natural language controlled devices may be implemented in an environment where the devices are configured to operate with multiple different users. The techniques described herein implement security requirements for a device configured to operate in an environment where the multiple different users may request functionality. The security requirements may be implemented based on the whether the functionality requested is personal and/or secured.