For the officer
Their account survives the typing
The notes go in the way they were written on site. What comes back is a draft in the right format, so the officer is not asked to explain their own report from somebody else's summary.
For security agencies · incident reports
The incident report that gets written twice
Your officer writes it at the site, in their own words. By the time a supervisor checks it, somebody has already retyped it into the client's format, and the version being checked is no longer the officer's account.
This is one workflow, taught inside Level 2 of OwnYourAI. An agency's own staff build one AI colleague that turns an officer's site notes into the incident report format that client already expects. The officer's account goes in as written, a draft comes out for a supervisor to check and sign, and the rules about what may be entered are written in Level 1 before anything is built.
Corporate pricing on request. Course funding application status: pending.
Where it is taughtLevel 2, one workflow
Built byyour own staff
Runs onyour process, not ours
Data ruleswritten in Level 1
Where the hour goes
One report, three sets of hands
The officer writes the account at the site, usually on paper, sometimes on a phone. Back at the office somebody types it into the client's report format, because the client wants their own headings, their own reference numbers and their own wording. A supervisor then checks the typed version, which is where the question usually starts: is this still what the officer said?
Then the client asks one detail six weeks later, and somebody goes back to the original to find it. None of this is a discipline problem. It is what happens when an agency grows from five people to thirty and no back office grows with it.
Meanwhile an officer finds a public AI tool that tidies a report in ten seconds, and it works. It is also the exact moment a client's incident details leave the agency's control with nobody able to say where they went. A ban that nobody can enforce is not a control: the fix is a written rule about what may be entered, plus a tool your team is allowed to use and knows how to use safely.
What changes, and for whom
Three people, three different changes in the week
For the supervisor
The check happens once, on the right version
The supervisor is checking the officer's words against the client's format, which is the job. Retyping is no longer part of the job, and the report is signed once.
For the director
The rule about client data exists in writing
Level 1 puts the agency's own rule on paper: what may go into a tool, what never does, and who checks. That is usually the thing directors wanted most and could not find a week to write.
These are descriptions of the work the training addresses. They are not client results and not a promised saving: the number belongs to your incidents and your client's format, and it is measured during your own level on your own reports.
What we measure on this one task
| What we look at | How it is measured | Who keeps the measurement |
|---|---|---|
| Note to signed report | The same report timed before and after, on your own incidents | Your company |
| Corrections needed | Counted on the supervisor's checks over a set number of reports | Your company |
| Format errors raised by the client | Logged against the client's own format requirements | Your company |
| What the tool must never see | The do-not-enter list agreed by the people who handle the documents | Your company |
Read this before you book
Four things this is not
- It does not decide anything. A person checks and signs every report that goes out. The tool drafts; the agency signs.
- It does not change the client's format or the agency's rules. Those stay exactly as they are — the build is wrapped around them.
- It is not a promise of fewer staff. What changes is where the hours go. The headcount decision stays yours.
- It is not for every agency. If a client contract forbids any third-party tool for their incident details, we say so at the assessment instead of selling around it, and we look for a different workflow.
What we are selling you here
Result, measurement, timing and remedy. Nothing else on this page matters as much as this block.
- The result
- One incident-report workflow handled by an AI colleague your own staff built, on the agency's own process, with a written rule about what may be entered and a named person who checks before anything is signed.
- How it is measured
- Time from note to a signed report, corrections per report, and format errors, counted on a set number of your own real incidents before and after — plus the do-not-enter list agreed before the build started.
- When it lands
- The draft runs on real reports inside the Level 2 build window agreed at enrolment, and is demonstrated on your own report, on inputs supplied fresh, at the Level 2 assessment.
- If it does not land
- If it is not running on real agency reports by the end of that window, we stay on with your team at no extra cost until it is, or that level is not charged. When your team carries the risk of a new way of working, sharing that risk is only fair.
Incident-report questions
What operations managers ask us
Does the AI write the incident report, or does the officer?
The officer writes the account. The AI colleague turns that account into the format the client already expects, and a supervisor checks and signs it. No report reaches a client on the tool's authority.
What stops client details leaking into a public AI tool?
A written rule, and a tool your team is allowed to use. Level 1 writes the do-not-enter list with the people who handle the documents; Level 2 builds the colleague they then use instead. Banning tools without either of those leaves the risk exactly where it is.
Our client contract forbids third-party tools. Can we still do this?
Then this particular workflow is not for you, and we will say so at the assessment rather than sell around it. There is usually a second task in the same agency that has no such restriction, and that is where we would start.
Do we need an IT person for this?
No. It is built for the people who already write the reports. If a step is technical, it is shown step by step, and your team can ask inside the scheduled touchpoints rather than waiting for a help desk.
Which level is this, and what does it cost?
It is a Level 2 use case: one workflow, built by your own staff. Corporate pricing is on request and quoted after the build assessment, because the number of seats and the starting level decide the work. Our course funding application status is pending, so we quote no subsidy and promise none.
Next step
Bring one week of your reports to the call
We will tell you on the call whether this is a fair target for the training, and what it would cost for your team.
Corporate pricing on request. Course funding application status: pending.