Can AI review a contract against your legal team's own criteria?
It can when the playbook is part of the system rather than a prompt. The agent flags where a draft departs from the company's standard position and names the clause it breaks; a lawyer confirms or overrides, and that decision feeds back in.
Description
An agent that reads a draft clause by clause against the company's negotiating playbook and ranks the deviations by risk, each pointing at the position it contradicts.
The problem
Legal reviews the high-value contracts. The long tail (renewals, suppliers, standard terms) goes through with a generic checklist or with nobody looking, and the playbook that should govern it sits in a document nobody opens at the moment of signing, so the criteria drift.
How we solve it
The playbook, the approved precedents and the standard positions live in RenBase, so the agent compares against what this company accepts, not what contracts look like elsewhere. Every flag cites the internal clause it comes from, which is what makes it arguable. RenLayer logs each review, so a year later you can see the decision and the basis for it.
What it returns
The long tail gets reviewed at the same standard as the headline deals, and the criteria stop drifting. Legal spends its time on the deviations, not the reading.
KPIs to track
What to measure to know whether this is working. We define the metrics with you before starting, chosen so that it is hard to look good without having done well.
- Share of contract volume reviewed
- Measured over the long tail, not the headline deals, which were already being reviewed.
- Precision and recall against the lawyer
- Flags the lawyer confirms, and deviations the lawyer catches that the agent missed. Recall is what decides whether this can be trusted.
- Deviations caught before signature
- Against those found afterwards. That migration is what you are measuring.
- Review time per contract
- End to end, with the lawyer's confirmation included.
Contact
Does your case look like one of these? That is the best way to start the conversation.
Talk to us