​
Overview

This document outlines repository classes used to fetch data from MongoDB in a structured and consistent manner. Each class method follows a standard pattern — first call initialization() with required arguments, then invoke the respective _fetch_... method.
These repository classes are part of the internal infrastructure and are meant for low-level data access within the Weam AI system.

​
CompanyRepository

​
_fetch_company_data

Description:
This function retrieves a company’s core metadata from MongoDB using its _id.

​
Usage

Call the initialization() method with:
  • company_id: str
  • collection_name: str
repo = CompanyRepository()
repo.initialization(company_id="123456", collection_name="companies")
company_data = repo._fetch_company_data()

​
ChatSessionRepository

​
_fetch_chat_session_model_data

Description:
Retrieves a Chat session’s data from MongoDB using _id.

​
Usage

  • company_id: str
  • collection_name: str
repo = ChatSessionRepository()
repo.initialization(company_id="123456", collection_name="chatsessions")
session_data = repo._fetch_chat_session_model_data()

​
LLMModelRepository

​
_fetch_company_model_data

Description:
Retrieves company model data using an API key identifier. It’s a core method for accessing LLM configurations across the codebase.

​
Usage

  • api_key_id: str
  • collection_name: str
repo = LLMModelRepository()
repo.initialization(api_key_id="apikey-abc", collection_name="llm-models")
llm_config = repo._fetch_company_model_data()

​
EmbeddingModelRepository

​
_fetch_company_model_data

Description:
Fetches embedding model configurations including model dimensions and other vector settings from MongoDB.

​
Usage

  • api_key_id: str
  • collection_name: str
repo = EmbeddingModelRepository()
repo.initialization(api_key_id="apikey-abc", collection_name="embedding-models")
embedding_config = repo._fetch_company_model_data()

​
PineconeRepository

​
_fetch_pinecone_company_data

Description:
Fetches Pinecone vector DB configurations including index name, environment, and other Pinecone-specific metadata.

​
Usage

  • api_key_id: str
  • collection_name: str
  • **kwargs: Optional custom query filters
repo = PineconeRepository()
repo.initialization(api_key_id="apikey-abc", collection_name="pinecone-configs")
pinecone_data = repo._fetch_pinecone_company_data(environment="us-west1")
Ensure that the Pinecone environment and index are valid, especially when passing custom **kwargs in production.

​
FileRepository

​
_fetch_file_model_data

Description:
Retrieves file-related metadata from MongoDB, used throughout the application for handling user or system files.

​
Usage

  • file_id: str
  • collection_name: str
repo = FileRepository()
repo.initialization(file_id="file-123", collection_name="files")
file_data = repo._fetch_file_model_data()

​
ChatMemberRepository

​
_fetch_chat_member_model_data

Description:
Fetches chat participant data used to manage individual members of a chat session.

​
Usage

  • chat_session_id: str
  • collection_name: str
repo = ChatMemberRepository()
repo.initialization(chat_session_id="chat-789", collection_name="chat-members")
member_data = repo._fetch_chat_member_model_data()

​
Developer Notes

  • All _fetch_... functions follow the same convention: call initialization() first, then call the method.
  • Errors are not automatically handled — consider wrapping calls in try/except blocks.
  • Data is returned in native Python dictionary format, directly from MongoDB.
This document is intended for backend developers working with internal data repositories. For abstraction-layer APIs or services, refer to the API reference.