It is generated by Claude, Anthropic's AI, from your own lesson notes and the pupil's current DVSA skill ratings. It is not a template with names swapped in — it reads what you actually wrote last time.
The morning
Know exactly what to teach before you pull up.
You have thirty pupils. You saw this one three weeks ago. You know there was something about roundabout lane discipline, or was that the other Maya? So you do what everyone does — you ask them how they got on last time, and take the first ten minutes of a paid hour working out where you are.
Written from your notes, not from a template
The briefing is generated by Claude from two things you already have: the notes you wrote after the last lesson, and the pupil's current ratings across the DVSA framework. That matters, because a template can only tell you what a lesson seven usually looks like. This tells you what happened to this pupil, in your car, three weeks ago.
It reads like a note from a colleague who taught them last — because functionally, that is what it is standing in for.
Four sentences. Ten seconds.
Length is a feature. A briefing you read in the ten seconds between parking and knocking gets read every time; a page of analysis gets read twice and then ignored. So it is capped at three or four sentences: what you covered, what they struggled with, what to focus on today. If you want the full picture it is one tap away in their progress record — but the briefing itself stays short enough to actually use.
Your data stays yours
Your pupils' information is sent securely to generate each briefing and is never used to train any AI model. Anthropic is a named sub-processor in our privacy policy, alongside Stripe, Neon, Vercel and Resend.
How it works
- 1
Rate the lesson at the kerb
A couple of taps against the DVSA skills you worked on, plus a line of notes if you want. Thirty seconds while they are getting their bag.
- 2
The next briefing writes itself
Generated fresh for the next lesson, drawing on what you just recorded.
- 3
Read it on the way
It is on the lesson card in your Today view, so it is there when you arrive rather than somewhere you have to go looking.
Questions about AI briefings
No. Your data is sent securely to generate each briefing and is not used to train any AI model. Anthropic is listed as a sub-processor in our privacy policy.
Related: Privacy policy
Three or four sentences. It is deliberately short — it exists to be read in the ten seconds before you knock on the door, not to be a report. Anything longer would not get read.
It is a prompt, not an instruction. It summarises what you recorded, so if your notes were thin the briefing will be thin. You teach the lesson you think the pupil needs — the briefing is there so you start from what happened last time rather than from memory.
The briefing can only work from what you give it, so sparse notes make a sparser briefing. The DVSA ratings do a lot of the lifting on their own, though — even with no written notes, it can see which competencies moved last time and which did not.
That is where it earns its keep. The briefing draws on whatever was recorded, whoever recorded it, so a pupil handed between instructors does not have to explain their own history at the start of every lesson.
Stop starting lessons from memory
Free until 31 March 2027. No card, no contract, cancel any time.
Get started free →Related
- Progress tracking → — the ratings the briefing is written from
- Scheduling → — where the lesson card lives
- How progress tracking works → — the longer walkthrough