Loading…
Use Case
We build AI contract review systems that read incoming contracts, extract every clause, compare against your playbook, and produce an annotated summary with flagged deviations, before a lawyer opens the document. Reviewers confirm flags and negotiate, rather than reading from scratch.
A contract that currently takes two hours of associate time to first-pass review takes five minutes with AI handling the initial pass. The same work is done; the lawyer spends their time on the parts that require judgment.
Tell us about your contract review workflow.
Six components that handle the full first-pass review workflow — from document ingestion through to a pre-organised reviewer packet.
Accepts PDF and Word (.docx) files. Handles varying formatting, tracked changes, headers, footers, and defined terms sections without losing structural context.
Your standard clause positions (what you accept, what you reject, and what requires negotiation) are encoded as your review playbook. The AI flags against your standards, not generic best practice.
Each clause is identified, labelled by type (indemnification, limitation of liability, IP ownership, governing law, etc.), and extracted into a structured review record.
Clauses that deviate from your playbook are flagged as high, medium, or low severity. High-severity deviations (e.g., uncapped liability) are highlighted immediately; low-severity issues are noted but don't block the workflow.
Each flagged clause is shown side by side with your playbook standard. Reviewers see exactly what changed and why it matters, no hunting through the contract to understand the deviation.
The system routes to the correct reviewer based on contract type or deviation severity, with a pre-populated summary of all flags and suggested redline comments attached.
From contract upload to reviewer packet, the full sequence.
Contract uploaded
The counterparty contract arrives via upload portal or email. The system ingests the file and begins processing, no manual routing to a legal queue required.
Clause extraction
Every clause is extracted and tagged by type. The extraction understands cross-references, defined terms, and exhibit references, the full document structure is mapped.
Compare against playbook
Each extracted clause is compared to your playbook position. The system identifies deviations, missing required clauses, and non-standard language.
Flag deviations with severity
Deviations are scored by severity: high (requires negotiation before signing), medium (flag for review), and low (note for awareness). The severity definitions are yours to configure.
Generate annotated summary
A review packet is generated: a clause-by-clause summary with flags, deviation descriptions, and suggested redline language for each non-standard clause.
Route to reviewer
The annotated summary routes to the assigned reviewer. High-severity deviations trigger an immediate Slack alert. The reviewer opens a pre-organised review rather than reading the contract from scratch.
The right fit is a legal or procurement team with consistent contract volume and a defined standard position.
Most of those contracts have the same three issues. The AI handles initial review and flags the issues so legal can focus on negotiation rather than reading.
Vendor contracts that sit in the queue for two weeks waiting for legal review delay procurement cycles. The AI completes the initial review in minutes so legal only needs to confirm the flagged items.
Volume contract review work that currently takes a junior associate days can be done in hours with AI handling the first pass. The associate reviews flags rather than reading every clause.
We’d rather decline than take a project that won’t deliver value.
One-off contract reviews without a repeatable playbook
The system flags deviations against your defined standards. If you don't have a standard position on key clauses, there's nothing to flag against. The first step is defining the playbook. If you don't have one yet, that's the conversation to start with.
Teams without a defined position on key clauses
If your legal team evaluates each clause on a case-by-case basis without a standard, the AI cannot consistently identify what's non-standard. The system works with teams that have a playbook (however informal) that we can encode.
Tell us which contract types you review most frequently, your approximate monthly volume, and whether you have a documented playbook. We’ll reply within one business day.