# Entity database

Kernel's entity database pairs traditional structured data with *entity memory:* semi-structured context, timelines, and notes drawn from raw, multi-source unstructured data (websites, PDFs, job posts, company filings, and more). 

This grounds every entity record in primary sources and lets users go beyond Kernel's core fields to extract the exact data points they care about, at a fraction of the cost of doing it from scratch.

#### Sources

Kernel draws on a variety of sources to populate coverage and context

1. **Web crawling:** Kernel's web crawler, e.g., crawling Unilever's website for all brands
2. **LinkedIn**: Kernel maintains a database of 70M LinkedIn account profiles, which is updated monthly
3. **Company regsitries:** Company registries and public databases - full list coming soon