Abstract: Techniques for building and maintaining model webs are described. In some examples, a model web is built by selecting models for the model web from one or more available model types based on at least one or more of availability, tensor information, and compute type, instantiating synapses between the selected models to form the model web and updating information regarding availability of the selected models of the model web to indicate being in use.

Abstract: In some examples, a method includes receiving a request to load a first machine learning model on an edge device, wherein the request includes an external identifier of the first machine learning model. The method further includes loading at least one instance of the first machine learning model onto the edge device, wherein the first machine learning model is loaded into a model pool having at least a second machine learning model using the same external identifier as the first machine learning model.

Abstract: This disclosure describes a method for segmenting large MQTT messages that exceed a maximum transport unit (MTU) for MQTT messages into smaller segments and then reassembling the segments into the original MQTT message. More particularly, the techniques and architecture provide two new fields within an existing MQTT message. In configurations, the two new fields may include a universally unique identifier (UUID) and a segment number. These two new fields enable segmentation of arbitrarily large messages at a publisher side of the MQTT message and reassembly of the segments of the large messages at the subscriber side so that the message size limitation may be overcome. This can allow for larger MQTT messages to be transmitted without the increased cost of larger maximum transport units.