extract-entities

To build a knowledge graph from emails and calendar events, you process synced files to automatically extract people, organizations, and topics. The system identifies key entities and links them together into a structured network.

Yes, you can generate Obsidian-compatible markdown notes from meeting transcripts. Processing extracts attendees, decisions, and action items, automatically creating and updating linked notes for each person and project discussed.

Yes, Node.js is a required dependency to run this entity extraction process. You must also configure your user identity and ensure your input files match the specific formats expected for processing.

The best way to structure ad-hoc documents for entity extraction is to place them in the specified input locations configured for processing. The system parses these documents alongside synced emails to identify projects and topics.

State tracking works by monitoring processing history to identify only new or updated calendar files. This efficient approach prevents reprocessing unchanged data, ensuring your knowledge graph is updated without redundant computations.

A limitation of building a knowledge graph from raw data is that extraction requires specific input file formats and user identity configuration. Without proper formatting, the system cannot accurately parse emails or documents to map relationships.