Project overview
I built review views where a reviewer can open the exact PDF page beside an extracted table. The portal shows row tags, pivoted comparisons and the table’s JSON. Reviewers can filter by parser and trace a discrepancy back to the source filing.
The engineering problem
Reviewers need enough context to assess extracted tables without repeatedly searching through the source document.
My contribution
- Built side-by-side PDF and extracted-table comparison views.
- Added links to the exact source PDF page.
- Implemented pivoted comparisons, parser filters, and JSON display.
- Added row-tag rendering in review views.
Built side-by-side PDF and extracted-table comparison views.How the system works
- Choose a parsed filing and its extracted tables.
- Display the source page beside its structured output.
- Expose table metadata and row tags for inspection.
- Keep source-page navigation close to the review workflow.
Tools & Technologies
PHP · MySQL · JavaScript · HTML/CSS
Delivered capabilities
Source-page navigation
Side-by-side comparison
Structured output review
Table metadata visibility
Project illustrations use sample data.
