I am also interested in your open-source-first approach, and am a bit confused from the landing page. Could you explain what exactly your product does? Is it an LLM-enhanced document parser?