US Patent 11281673 Data pipeline for scalable analytics and management

Systems and methods are disclosed for performing computations on data at an intelligent data pipe en route to a data store. In one embodiment, a method is disclosed, comprising: receiving metadata regarding a data stream from a data source; performing an analysis of the metadata at a service orchestrator; creating at least one container instance based on the analysis; streaming the data stream from the data source to a data sink via the at least one container; and processing the data stream as it passes through the at least one container instance, thereby enabling application-aware processing of data streams in real time prior to arrival at the data store.