Member Junction
    Preparing search index...

    Generates entity groups using LLM-based semantic analysis

    This class replaces the deterministic hub-and-spoke algorithm with an intelligent approach that understands business context and generates meaningful entity combinations for query generation.

    Index

    Constructors

    Methods

    Constructors

    Methods

    • Generate semantically meaningful entity groups using LLM analysis

      Note: The LLM determines the optimal number and size of entity groups based on business domain understanding. Makes separate LLM calls for each schema to avoid mixing entities from different databases.

      Parameters

      • entities: EntityInfo[]

        All entities to analyze

      • contextUser: UserInfo

        User context for server-side operations

      Returns Promise<EntityGroup[]>

      Array of validated entity groups with business context