narrative-extractor

To extract structured data from unstructured narratives into a Neo4j JSON schema, provide your caregiver notes or interview documents. The skill applies strict non-guessing rules and canonicalization to normalize entities and dates safely.

Yes, you can extract entities from attached PDF and DOCX files for Neo4j ingestion. The skill processes attached documents including docx, pdf, xlsx, and txt formats, converting unstructured text into structured JSON data.

The best way to normalize name variants and resolve entities from care planning narratives is using strict extraction rules. The skill canonicalizes name variants and performs entity resolution against existing Neo4j data to ensure consistency.

The extraction process handles implicit insights by prioritizing them while enforcing strict non-guessing rules. It avoids fabricating information by only capturing care preferences and support logs explicitly mentioned in the source narratives.

A limitation when converting unstructured interview notes is the strict non-guessing rule, meaning the extraction will not fabricate missing information. It relies entirely on the explicit and implicit insights present in the provided text or attached files.

No, you do not need existing Neo4j data to extract structured data, but having it enables entity resolution. The skill can resolve canonicalized name variants against existing Neo4j records to maintain connected client relationships accurately.