extract

To extract specific facts from a note without introducing new data, use LLM-guided content derivation. This process grounds the output entirely in the source text, pulling targeted sections while ensuring no external information is added.

Yes, you can compress and summarize lengthy documents grounded only in their original text using LLM-guided transformation. This ensures the summarized content strictly reflects the input document without synthesizing information from other sources.

The best way to reformat text content into JSON or bullet points from a single source is through LLM-driven text reshaping. This transforms the document structure while guaranteeing the output remains completely grounded in the original input.

No, content derivation does not support cross-document synthesis for generating reports. It strictly processes a single Note at a time, ensuring the output is grounded entirely in that specific input and no cross-document analysis occurs.

You should not use single-note extraction when you need to synthesize information across multiple documents or introduce new data. This approach is strictly limited to transforming, compressing, or extracting from one source note at a time.