How it works
The review runs on a document of a contractual project, for example a supplier SaaS agreement returned by the counterparty. The agent reads it against the template the project was created from, the applicable clause variants and the playbooks written by your team.
Each finding is typed as divergent, missing or miscellaneous, comes with its reasoning, and is grouped as high, medium or low risk. If nothing is wrong, the report says that no mismatch and no missing clause were found.
From each finding you can ask the agent to propose a fix in the chat, or create an action assigned to a member of the project team. The action shows in the project activity feed.
What you get
- Comparison with your template and clause library
- Your own playbooks and fallback positions
- Items grouped as high, medium or low risk
- Fix in chat or assign an action to a teammate
Why teams use it
Legal starts from a ranked list of issues instead of a blank read.
Playbooks hold preferred and fallback positions, so sales can see what they may concede and where the red lines are.
The review also flags what should be there and is not.
Turn an issue into a proposed fix or a task for a colleague in one click.
Simple contracts get a first check without waiting for legal.
More in AI & extraction
Key data extracted, with the source clause
Every value linked to the clause and page it comes from.
Learn more →AI summary sheets
A structured summary for each contract, set up per agreement category.
Learn more →Chat with a contract
Ask questions about one contract and get sourced answers.
Learn more →Questions in plain language across your portfolio
Ask all your contracts a question at once.
Learn more →