As it indexes documents, Fusion invokes ML for clustering and classification, NLP functions like named entity recognition (NER), sentence detection, part of speech tagging, regex, and sentiment analysis with deep learning.
As data flows out at query time, Fusion uses ML again to interpret query intent to return the most relevant results. Other query-time AI identifies domain-specific entities (people, places, things, synonyms, misspellings, topics, phrases, etc.) with the semantic knowledge graph and understands their relationships within the searched content.