Investigate llms capabilities with structured PDF docs

この issue には説明がありません。

The issue title suggests exploring LLM capabilities for structured PDF documents, but the body is empty. Start by checking the repository's existing AI/ML modules, especially any document processing or computer vision components. Look for related issues or PRs to understand the context. Determine what 'structured PDF docs' means in this project—whether it's about parsing, extracting data, or integrating with the ERP system. The goal is to define a clear scope and propose an implementation approach.

索引モデルが issue の本文から書いたものです。