The contract queue is where routine paper eats the hours that belong to real matters, intake lives in untracked email, and generic AI is a confidentiality non-starter. An AI-native legal function reviews every inbound agreement against your own playbook the day it arrives, while nothing leaves the building without counsel.
Installed in about 20 min. Governed and audited from the first run.
Every stage of the motion is worked by the right kind of worker (an unattended play, a room of specialists for the hard agreements), all reading and writing one shared record, with counsel at the gates that matter.
You don't wire this together or choose an architecture. You describe the outcome; the workforce brings the right kind of worker to each part of the motion, and they all work from one shared record, under one set of rules.
The repeatable plays (intake an agreement, run the first pass against your playbook) go start to finish on their own, unattended, the moment a contract lands, day or night.
When an agreement is complex enough that liability, term and IP have to be weighed together against the playbook, a room of specialists assembles the review, each owning its part.
You don't operate a dashboard. You talk to one coworker that runs the intake desk, holds your standing instructions, and brings you only the clauses that need counsel.
Every worker reads and writes the same matters, playbook and documents, and every one obeys the same gates, budgets and audit. No worker is ever off the leash.
Confidentiality is enforced, not promised: access scoped by org role, a PII policy over what agents store, and nothing external released without counsel at the gate.
There's no integration project and no architecture to pick. You delegate the outcome, the workforce assembles the right workers, and every one of them lands under the same gates and audit.
Not a chatbot with a law degree it doesn't have: a workforce that does the first pass against your positions and hands counsel the judgment calls with the evidence attached. The week shifts like this (a worked scenario, not a benchmark):
Every inbound NDA and routine contract is read against your playbook (your standard positions, fallback clauses, walk-away terms) the day it arrives, and each deviation names the rule it breaks. The contract queue stops being where deals go to wait.
Every agreement that arrives becomes a matter row in a typed table: counterparty, type, status, owner, due date. Nothing lives in someone's email; nothing goes quiet; nothing gets rediscovered late.
Questions about a playbook rule or a precedent are answered from your own knowledge bases with the citation attached. Anything the documents can't ground escalates to counsel instead of being guessed at.
No redline sent, no signature, no filing without an approval at the gate: one single-use grant per release, and every access on the audit record. A defensible record, by construction.
Lex isn't a dashboard you operate, it's the coworker you delegate the first pass to. The night's reviews, the deviations found, the redlines that are ready: it arrives summarized, the clauses that need counsel are already queued with their cited rules, and each approval releases exactly one thing.
Every message, tool call and approval in this exchange lands in the audit trail. How the gates work →
A scenario: an in-house team with a two-page NDA playbook and a rule that nothing external moves without counsel. These are real screens, not mockups.
Every flag cites the playbook rule it deviates from; document answers come from your knowledge bases with citations, and gaps escalate to counsel instead of being guessed. Inbound contract text is untrusted input, so a prompt-injection shield screens it before it reaches a model. The control plane →
Redlines, signature requests, filings: every external release holds at the approval gate, enforced at the tool call, at runtime, so an unapproved send is a no-op. Each approval is single-use: one grant, one release, 24-hour expiry, and counsel can delegate approvals when out of office. How approvals work →
A PII policy with redaction governs what agents store; every document access and tool call lands in the audit trail, pinned to the config version that ran; and every run's spend rolls up against budgets with hard caps. When someone asks how a matter was handled, you show the record.
