Browse docs
Documents
Review tables
Extract structured data from multiple documents at once using AI-powered extraction grids
A Review table is an AI-powered extraction grid. You define the columns — each column is a question or data point you want extracted — add a set of documents, and the AI fills every cell. The result is a structured comparison across your entire document set.
This is 8rney's equivalent of the document review grids used by large-firm lawyers and tools like Harvey AI — available at the matter level without custom configuration.
Creating a Review table#
Open Review tables
Click Review in the sidebar, then New table. Give the table a name that describes the document set (e.g. "Target Co — NDA Portfolio Review").
Add documents
Click Add documents to pick from your library or matter. You can add up to dozens of documents to a single table.
Define columns
Click Add column and describe what you want extracted in plain language. Examples:
- "Governing law"
- "Termination clause — notice period"
- "Liability cap (as a multiple of contract value)"
- "Non-compete duration"
- "Is there a most-favoured nation clause?"
Run the extraction
Click Run to start the AI extraction. The table fills in real time — you can see results arriving cell by cell.
Review and refine
Click any cell to see the full source quote and AI reasoning. Re-run individual cells or entire columns if you want to adjust the extraction prompt.
Column templates#
Pre-built column templates are available for common review types:
| Template | Columns included |
|---|---|
| M&A — NDA review | Parties, governing law, confidentiality scope, term, exclusions, return of materials |
| Employment contracts | Role, jurisdiction, notice period, non-compete, IP assignment, severance |
| Real estate | Property description, term, rent, break clauses, assignment rights |
| Commercial agreements | Governing law, payment terms, liability cap, termination, dispute resolution |
Templates are starting points — add, remove, or rename columns to match your specific review.
Improving column prompts#
If a column is returning inconsistent or low-quality extractions, click the column menu and select Improve prompt. The AI rewrites the column prompt to be more precise based on the patterns it has seen in the document set.
Bulk actions#
- Select multiple cells and click Re-run to refresh those extractions.
- Select multiple cells and Export selection to download just those columns to CSV.
Column reordering#
Drag columns to reorder them. The column order is preserved in exports.
Querying the table#
Use the Ask field above the table to ask questions across the entire review. For example: "Which documents have a notice period longer than 30 days?" or "Which documents are missing a dispute resolution clause?". The answer cites specific cells.
Exporting#
| Format | Notes |
|---|---|
| CSV | Clean spreadsheet for further analysis in Excel or Sheets |
| DOCX | Formatted table with source document names and column headers |
| JSON | Structured data for integration with other systems |
Review table extractions are based on the document text available at the time of extraction. If you update a document in your library, re-run the relevant cells to refresh the results.
On this page