Abstract: A service to provide anomaly detection receives a request to back-test the service. The request includes information for accessing a dataset of historical data. The service executes workflows to ingest the data, train a plurality of machine learning models to perform anomaly detection, and detect anomalies in the dataset. A representation of the detect anomalies is generated and presented to a user. The service receives an indication to activate the service to provide ongoing anomaly detection services.

Abstract: Systems and methods are described for using a recurring event based scheduler to continuously monitor data to detect anomalies within the data. In some aspects, an anomaly detection schedule may be determined for monitoring time series data to detect anomalies based on a data ingestion interval. A plurality of anomaly detection events may be sequentially generated and stored in an event queue at times specified by the anomaly detection schedule. The anomaly detection events may then be processed sequentially from the event queue to trigger execution of a plurality of anomaly detection workflow tasks at the times specified by the anomaly detection schedule. In some cases, execution of individual anomaly detection workflow tasks causes individual portions of time series data to be obtained from a customer data source and processed by an anomaly detection model to detect anomalies in the time series data.

Abstract: Techniques for dynamically allocating storage space at fulfillment centers for sellers offering items are disclosed herein. In embodiments, information about available bins of a fulfillment center for storing inventory may be obtained. Sales performance factor information for a seller offering items may be received for items that are being stored at a fulfillment center. A capacity of bin numbers and bin sizes may be determined for the seller based on the sales performance factor information and the number of available bins. A threshold space assignment for the seller may be determined based on an inventory group associated with the offered items associated with the seller. Particular bin sizes and bin numbers may be identified at the fulfillment center for the particular seller to store the offered items based on the capacity and threshold space assignment.

Abstract: Techniques for automatically activating a count of device-independent functions to process a received request and other predicted traffic include determining a current count of active functions, a count of functions currently processing requests, a duration for execution of a function, an average time used by a function to process a request, and a current count of client devices accessing the system. A probabilistic value indicating the likelihood of additional traffic may also be determined based on characteristics of the request. These values are used to determine the available capacity for processing requests using the currently active functions and the predicted capacity to be used to process predicted additional traffic. A number of additional functions may be activated based on the difference between the available capacity and the predicted capacity.