LLM based classification

Classification module

Large Language Model based classification main handlers.

embedding.EmbeddingHandler	Handles embedding operations for the Chroma vector store.
llm.ClassificationLLM	Wraps the logic for using an LLM to classify respondent’s data
prompt.PromptTemplates	Class to store prompt templates for SOC and SIC classification tasks.

Response models

SocCandidate	Represents a candidate SOC code based on provided job title and description.
SocResponse	Represents a response model for SOC code assignment.
SicCandidate	Represents a candidate SIC code with associated information.
SicResponse	Represents a response model for SIC code assignment.
RagCandidate	Represents a candidate classification code with associated information.
RagResponse	Represents a response model for classification code assignment.

SIC Index Abstraction

Data models to represent Standard Industry Classiffication

SIC	Main class for SIC lookups.
SicCode	Standard Industrial Classification code.
SicNode	Tree data structure where the nodes hold all data associated with a given SIC.

SIC metadata

sic_meta_model.ClassificationMeta	Represents a classification meta model.
sicDB.SicMeta	SIC Meta data model class for SIC codes and their descriptions

Helpers

Config and Log utils

setup_logging	Set up console and file logging.
get_config	Fetch the configuration.
check_file_exists	Check if the file exists.