The challenge: notes that ate the billable day
The firm's consultants were spending hours each week writing up meeting notes - work that mattered but earned nothing, and that came out of the time they could have spent on client work. The notes were also inconsistent: each consultant captured different things, so the record depended on who happened to be in the room.
What we built
We built a documentation workflow around the meetings the consultants were already running, so the write-up happened on its own rather than as a separate evening job.
Record and transcribe
Meetings are recorded and transcribed, capturing the detail as it is said rather than from memory hours later.
AI summarisation
AI turns each transcript into a structured set of notes - decisions, actions, and follow-ups - in a consistent format every time, whoever ran the meeting.
Into the workflow
The notes drop into the firm's existing tools, so nobody has to copy them across or reformat them.
The result
Documentation time fell by about 80 per cent - roughly six hours a week back per consultant - and the notes got better, not worse. They now capture detail the manual write-ups used to miss, and they read the same whoever was in the meeting.
Why it worked
The work AI took over was the part the consultants least wanted to do and machines do well: capturing and structuring what was said. The judgement - what the meeting meant and what to do next - stayed with the people who bill for it.