Hundreds of contracts, one view
Due diligence at scale. Upload your document set, define your extraction questions, and get a structured comparison across every contract, instantly.
Ask once,
extract across every document.
AI spreadsheet a grid where every column is a question and every row is a document, filled by the agent.
Parallel extraction runs the same questions across hundreds of documents at once, not one after another.
Source linking every cell links back to the page and clause it came from, highlighted in the original.
Structured output export the finished grid to Excel or CSV, one column per question, ready for the data room.
Common questions
How does tabular review work?
You upload a set of documents and write your questions as columns. Every document becomes a row, and the agent fills in each cell by reading that document for that question. What you get back is a grid you can scan down a column to compare a hundred contracts on one point.
How many documents can I run at once?
Hundreds. The documents are processed in parallel rather than one after another, so the run takes about as long as the slowest document rather than the sum of all of them.
Can I trust a cell without opening the contract?
You can check it in one click. Every cell links back to the page and clause it came from, highlighted in the original document, so verifying an answer is a glance rather than a search. Cells the agent could not answer from the document are marked rather than guessed.
What kinds of questions work well as columns?
Anything with a findable answer in the document: governing law, notice period, lock-in, escalation, termination triggers, indemnity caps, assignment restrictions, change-of-control language. Questions that need judgement across documents are better asked of an agent than of a column.
Can I change the questions after a run?
Yes. Add a column and only the new question runs across the set. The existing answers are kept rather than recomputed. Removing a column does not touch the rest of the grid.
How do I get the results out?
Export the finished grid to Excel or CSV, one column per question, ready to drop into a data room or a diligence report. The source references travel with it.
Is this only for due diligence?
It is the obvious use, but any repeated question across a pile of documents fits: lease portfolios, employment contracts after an acquisition, vendor agreements before a policy change, or a compliance sweep across a set of licences.
