Use case · Law firms

Drafting and research assistance

First drafts from the firm's own precedent, with sources a lawyer can check.

The problem

Fee earners are already using AI for first drafts. Some of it is good, some of it is confidently wrong, and none of it is consistent between practices.

How we would run it

  1. STAGE 1

    Evidence

    A written picture of the current state, with its gaps named.

    Find out what people already do, including the tools nobody approved. That is the real starting point, whatever the policy says.

  2. STAGE 2

    Baseline

    A baseline measurement both sides agree on.

    Measure the drafting tasks that repeat: time, review effort, and how often the first draft survives to the second.

  3. STAGE 3

    Design

    A design your architecture and risk leads have approved.

    Anchor drafting in the firm's own precedent, design the review step, and set which tasks are in scope. Research assistance must cite, or it is not in scope.

  4. STAGE 4

    Ship

    Working software or a live process, in production.

    Release to one practice with worked examples and named owners, inside the approved tooling.

  5. STAGE 5

    Prove

    A result measured against the baseline, and an instrument you keep.

    Report against the baseline, including the tasks where it did not help, and use that to decide the next practice.

What you get

  • Tested prompts and workflows per drafting task
  • Drafting grounded in the firm's own precedent
  • Review step designed into each workflow, with an owner per practice
  • Before-and-after measures per task

What it needs from you

  • Precedent and know-how for the tasks in scope
  • The list of tasks that actually repeat
  • A practice group willing to go first

Where the human stays

Output is a first draft for a lawyer, not advice. Research assistance cites its sources, and anything uncited is treated as unverified.

How we would know it worked

  • Drafting time per task, against the baseline
  • Review effort and correction rate
  • Adoption inside the practice, four weeks on

Where it starts

Starts as a sprint

AI Governance Quick-Start

A firm-wide AI policy that respects privilege, and a list of what is approved.

Sprint details →

Grows into

AI Adoption & Enablement

Get AI into the work your fee earners actually do, with evidence for clients.

Program details →
Scope this as a sprint →
Start here

Start with one sprint

Two to three weeks, one approver, a deliverable you keep. Tell us the problem and we'll come back within one business day with a scope, a date and a fee.

Scope a sprint →