Systems and methods for generative ai powered use of hierarchical data structures

A system for data mapping, the system comprising:

• one or more memories comprising instructions stored thereon;
• and
• one or more processors configured to execute the instructions and perform a set of operations comprising:
• receiving the unstructured data including text data;
• generating an extraction prompt using the unstructured data;
• providing the extraction prompt to a generative artificial intelligence (GenAI) model
• receiving one or more extracted data parts from the GenAI model;
• for each data part of the one or more extracted data parts,
• generating a description prompt based on the each data part;
• providing the description prompt to the GenAI model;
• receiving a keyword description for the each data part from the GenAI model;
• accessing a hierarchical data model, the hierarchical data model including one or more sub-hierarchies, at least one of the one or more sub-hierarchies including a plurality of levels of structure elements;
• obtaining one or more highest-level structure elements in the hierarchical data model, each highest-level structure element of the one or more highest-level structure elements corresponding to at least one of the one or more sub-hierarchies;
• generating a search prompt using the unstructured data and the one or more highest-level structure elements;
• providing the search prompt to the GenAI model;
• receiving one or more identified highest-level structure elements from the GenAI model, the one or more identified highest-level structure elements corresponding to one or more identified sub-hierarchies, the one or more identified highest-level structure elements being a subset of the one or more highest-level structure elements;
• for each data part of the one or more extracted data parts, traversing each identified sub-hierarchy of the one or more identified sub-hierarchies to determine a list of relevant structure elements in the hierarchical data model based on a relevance scoring;
• and
• generating a mapping prompt based on the list of relevant structure elements and the unstructured data.

A non-transitory computer-readable storage medium having instructions for data mapping that, when executed by one or more processors, cause the one or more processors to perform operations comprising:

• receiving the unstructured data including text data;
• generating an extraction prompt using the unstructured data;
• providing the extraction prompt to a generative artificial intelligence (GenAI) model receiving one or more extracted data parts from the GenAI model;
• for each data part of the one or more extracted data parts,
• generating a description prompt based on the each data part;
• providing the description prompt to the GenAI model;
• receiving a keyword description for the each data part from the GenAI model;
• accessing a hierarchical data model, the hierarchical data model including one or more sub-hierarchies, at least one of the one or more sub-hierarchies including a plurality of levels of structure elements;
• obtaining one or more highest-level structure elements in the hierarchical data model, each highest-level structure element of the one or more highest-level structure elements corresponding to at least one of the one or more sub-hierarchies;
• generating a search prompt using the unstructured data and the one or more highest-level structure elements;
• providing the search prompt to the GenAI model;
• receiving one or more identified highest-level structure elements from the GenAI model, the one or more identified highest-level structure elements corresponding to one or more identified sub-hierarchies, the one or more identified highest-level structure elements being a subset of the one or more highest-level structure elements;
• for each data part of the one or more extracted data parts, traversing each identified sub-hierarchy of the one or more identified sub-hierarchies to determine a list of relevant structure elements in the hierarchical data model based on a relevance scoring;
• and
• generating a mapping prompt based on the list of relevant structure elements and the unstructured data.