data-research

To extract structured data from emails and web pages, this Skill uses a seven-phase pipeline to search, classify, and extract information, updating canonical tracker pages with backlinks for organized reporting.

The best way to deduplicate extracted data is using the pipeline's built-in deduplication phase, which processes extracted archives to ensure only unique structured records update your canonical tracker pages.

Yes, you need YAML recipes to extract investor updates and expense data. The Skill includes built-in recipes for these use cases, but also supports custom YAML recipes stored locally under your recipes directory.

The data extraction pipeline classifies and archives scattered research data by progressing through define, search, classify, and extract phases, scaffolding a local workspace to audit and structure the collected information.

Yes, you can define custom extraction schemas and sources for company updates research by creating custom YAML recipes, tailoring the seven-phase pipeline to fit your specific data governance and reporting needs.