dotData Stream is a containerized ML model that provides real-time prediction capabilities. Once a model has been built using the dotData Enterprise, it can be exported as a docker image with an embedded model and provide prediction service independent of dotData Enterprise. The only prerequisite to Stream is the ability to launch docker.
Overview
dotData Stream™
Real-time Prediction
dotData Stream is ideal for real-time analytics in resource constrained environments and IoT edge deployments. It is good for a single prediction or mini-batch up to one thousand predictions. It allows customers to scale throughput linearly by running multiple instances.
Ease of Use
dotData Stream is very simple to use and run offering a standardized approach using docker. It is easy to deploy in the cloud or at the edge. Stream is flexible, scalable and compatible with AWS SageMaker.
Ideal For Low Latency Applications
dotData Stream is very simple to use and run offering a standardized approach using docker. It is easy to deploy in the cloud or at the edge. Stream is flexible, scalable and compatible with AWS SageMaker.
Operationalization
dotData Stream is lightweight, requires low memory footprint and can be deployed on a single board computer like Raspberry Pi. Container orchestration can be performed using Kubernetes or Docker swarm. Stream can also run on Azure using Kubernetes.