PRISM facilitates AI/ML/DL operationalization as the connective tissue:
acquire data and make ready for training and/or inference;
fine tune the compute continuum with insight (data-flow taps) for human investigation and visibility into decision processes;
provenance and audit trail to understand how interactions between data and model present results (likelihoods);
understanding the complexity of interconnected processing from edge to cloud, and back again.
Behavior-as-a-Service “BaaS” software agents are different in that they are fundamental, foundationally atomic, computational units supporting inference, decision making, knowledge discovery, and training. PRISM is deployed using standard DevOps/DevSecOps toolchains.