AI Legal Triage: Tools for Faster Intake
Centralized intake is the strongest predictor of a mature legal team. What to capture for each request type, how to keep intake short, and AI that reads email and Slack.
Last updated November 2026
Legal triage is only be as good as the information it starts with. A request that arrives as one line in Slack, with no counterparty, no deadline, and no document, can't be prioritized or routed until someone writes back and waits for additional context. A finance respondent in Streamline's 2026 Legal Friction benchmark described what happens next from the requester's side: "They don't ask questions if there is something missing. They just reject and have you start over."
Faster triage starts at intake. Streamline's State of Legal Efficiency report, built from 79 legal teams' benchmark submissions, found that centralized intake with dynamic forms was the strongest single predictor of a mature legal operation, and scattered intake channels the most defining trait of low performers.
AI helps from both directions: it pulls details out of free-text requests, and asks a short follow-up when something is missing. Below is what to capture for each common request type.
The intake checklist, by request type
NDA
Counterparty name, mutual or one-way, purpose of the discussion, and whose paper it's on. Four fields, often all readable from the email.
Vendor or customer contract
Counterparty and contract type, value and term, signing deadline and the reason for it, whether personal data or a security review is involved, the draft, and any prior agreement with the same party.
Marketing review
The asset, where and when it will run, any claims, comparisons, endorsements or contests, and any use of third-party names, logos or images.
Employment question
Type of issue (hiring, leave, performance, separation), the employee's location, timing, and a note to share personal details only with the assigned lawyer.
General question
The question in the requester's own words, the business decision it affects, and when an answer is needed.
Keeping intake short enough to use
Every extra question is a reason to email a lawyer directly. Pre-fill whatever the message already says. Ask conditional questions only when they apply, such as personal data only for contract types that suggest it. Pull the requester's name, team and region from their profile. Ask follow-ups in the same Slack or email thread instead of sending a form. And review the fields every quarter: if nobody uses a field to prioritize or route, cut it.
From intake to triage
Once intake is complete, triage can run straight away: priority from the deadline and value, path from the request type, owner from your routing rules. Our guides to triage and routing cover those steps.
One team's result
With Streamline AI, Lighthouse had its intake forms and workflows running within a month. Within two months of rollout, the team cut time spent tracking down information by 40 to 50%.
Tools for faster intake
Streamline AI takes requests from email, Slack, Teams, and forms, reads what it can, asks short follow-ups for the rest, and hands a complete request to triage and routing. Each request type can have its own questions. Intake fields are yours to design; Streamline won't know which facts your lawyers need until you tell it. Book a demo.
Tonkean builds a no-code intake front door shared by legal, procurement and finance, with configuration needed for legal-specific questions. Josef offers guided questionnaires and document generation; check how it captures requests that arrive by email or Slack. Onit includes matter intake within a broader suite of contract, e-billing and spend tools, with a longer implementation than a focused intake tool.
Where to start
Take your five most common request types and write the minimum fields for each. Cut anything you wouldn't use to prioritize or route. Then copy three real requests from your inbox and check whether those fields would have been filled.
.avif)