Cross-Source Matches
Likely matches between records extracted from different source documents. Confirm to link them into a web of connected information — records stay separate, each preserving its own source citation.
Find Cross-Source Matches
Scans all extracted individuals across every source (yearbooks, history books, directories, stories) and creates reviewable match suggestions for the same person appearing in different sources. Temporal-aware — won't suggest impossible matches (e.g., a 1950 senior matching a 1921 directory entry). Does NOT merge or modify records.
Auto-Link People to City Directory
Matches Individuals to directory entries by exact name (handles "First Last" vs "Last First" format). Populates directory source citations on each person's page — no manual review needed.
Auto-Build Family Groups
Groups Individuals by shared surname and creates Family records for groups with explicit relationships (spouses, parents, children) in the history books. Each family page shows a "probable connections" disclaimer.
Extract People from Historic Sites
Uses AI to read each historic site's text and identify people with direct relationships (owners, builders, founders, architects, etc.). Matches against existing Individual records and creates reviewable link suggestions.
Extract Businesses from Historic Sites
Uses AI to read each historic site's text and identify businesses with direct relationships (occupants, builders, owners, operators, publishers, etc.). Matches against existing Business records and City Directory entries, creating reviewable link suggestions and auto-connecting to directory appearances.
Extract People & Businesses from Stories
Uses AI to read each story's content and identify people and businesses significantly mentioned. Matches against existing Individual and Business records, creates reviewable link suggestions, and auto-creates new records for unmatched names.
Extract People & Businesses from Timeline Events
Uses AI to read each timeline event's title and description, identifying people and businesses significantly mentioned. Matches against existing records, creates reviewable link suggestions, and auto-creates new records for unmatched names. When an event is linked to a historic site, extracted connections are also suggested on that site's page.