Shruti is an architecture for massively parallel, scalable reflexive inference. The importance of Shruti for the CognitiveWeb lies in its ability to rapidly fuse together fuzzy or probabilistic information with a long-term knowledge base and produce explanations, predications, and plans. Shruti is used to combine evidence and expected utility in the evidentiary graphs derived from argument models. Shruti is also used to localize the sources of uncertainty within those argument models. This information is used to drive the facilitated critical thinking and conflict resolution processes on the CognitiveWeb.

See the Shruti link off of Lokendra Shastri's home page for more information.

An open-source shruti architecture is now available as a CognitiveWeb project. There is also a wiki page for this project.