Case studies

Proof, measured against a baseline.

These clients are anonymised because they have not agreed to be named. The numbers are not. Each one comes from a metric the business was already tracking before we started, measured the same way after. Where we cannot show you the working, we do not publish the number.

Conveyancing and property law

45%

faster intake to instruction

Intake to instruction, cut by nearly half

The situation

A growing conveyancing firm was losing work at the front door. Referrals arrived through several channels and were routed by hand, status updates required someone to chase them, and referral partners had no way to see where a matter stood without picking up the phone.

What we built

A connected intake platform that routes referrals automatically, gives every stakeholder a live view of matter status, and issues milestone updates without anyone in the firm triggering them. Every action is logged for compliance and the data stays hosted onshore.

What changed

  • Referral partners submit and track matters themselves, so the phone tag stopped
  • Milestone updates reach all parties without staff intervention
  • A full audit trail covers the intake process end to end

How it was measured

Intake to instruction time was already tracked by the firm as an operating KPI before the work started, which is what makes the comparison meaningful. The platform was built with that measurement in it, so the before and after come from the same definition and the same tracking, not from two different methods.

Custom platform Automated routing Live dashboards Onshore hosting

Legal, land sale contracts

1-2 hrs to ~10 min

per contract reviewed

Contract review from one to two hours down to about ten minutes

The situation

Senior reviewers were reading land sale agreements end to end and filling in a review template by hand. It was accurate work, and it was consuming the most expensive hours in the firm on a task that was mostly extraction rather than judgement.

What we built

A private AI agent that reads each contract, extracts the key clauses, flags the risk items and pre-fills the standard review template. The reviewer validates and decides rather than starting from a blank form. Documents never leave the firm's own environment.

What changed

  • Reviewers validate a pre-filled template instead of building one from scratch
  • Key clauses and risk flags are extracted consistently on every contract
  • Documents stay inside the firm's private environment throughout

How it was measured

Review time per contract was captured through the team's existing time tracking, which is how the one to two hour baseline was known in the first place. The agent's output flows through the same review step, so the after figure is measured the same way as the before figure.

Private AI Document analysis Template automation Onshore infrastructure

Why the method matters

A number is only worth what you can trace it to.

Most AI results you read about are estimates written after the fact, compared against a baseline nobody recorded. Both results above come from metrics the business was already tracking, because we build measurement into the system rather than bolting a report on at the end.

Baseline agreed first

We write down what today costs in hours, turnaround or spend before anything is built.

Same definition after

The after number uses the same metric and the same tracking, so the comparison holds.

Yours to check

The measurement runs in your systems, so you can audit the result without taking our word for it.

Your turn

What would we measure in your business?

Thirty minutes with Liam or Nick to find the work with the clearest return and agree what the baseline should be. If the numbers would not justify it, we will say so on the call.

Book a scoping call Try the 2 minute assessment