This is a public prototype. Retrieval runs only over a sanitized corpus stored locally on this machine.
Try these queries:
"fraud detection rules"
"what is canonical normalization"
"explain equivalence classes in DSL"
"example of IF x THEN y rule"
"model transformation function"
What this demo does
This prototype shows a different kind of search. Instead of matching exact words,
it transforms your query into a canonical form—a standardized version of
meaning—before searching.
That means it can find relevant results even when:
the wording is different
the structure changes
or the same idea is expressed in multiple ways
In simple terms: It searches by meaning, not just by text.