Skip to main content

Safeguarding assist

Helps you write the expectations and concerns fields on your trust's own safeguarding referral form.

It does not submit anything anywhere. It produces wording; the referral goes through your service's process exactly as it always did.

How it works​

Unlike the other AI tools, this one is a conversation. It asks one question at a time, drawing out what concerned you and why, then drafts the fields from your answers.

The interview format is the point. Safeguarding referrals fail on vagueness — "concerns about home environment" tells the receiving team nothing — and the questions push for the specific, observed detail that makes a referral actionable.

Privacy​

This tool applies the strictest handling in MedJot.

  • Redaction runs on every field before anything is sent, on both the device and the server.
  • It never outputs a full name. Not the patient's, not a family member's, not a professional's.
  • As with every AI tool, the text sent and the response returned are not stored.

You will be adding names into your trust's form when you file it. That happens in your trust's system, not here.

See What gets sent.

What to check before filing​

  • The concern is described, not characterised. What you saw and heard, not a conclusion.
  • Nothing has been softened. The draft's register is professional, and professional wording can flatten something that needs to sound as serious as it is.
  • Nothing has been added. As everywhere, the tool works from what you told it — check that it has not stretched an inference into an observation.

Why it cannot be queued​

If the request fails for lack of signal, this tool is not offered for queueing. It is a conversation, and a queued turn would arrive with no conversation around it. Reconnect and start the interview.

The referral is yours​

MedJot drafts wording. The judgement that a referral is warranted, its urgency, and everything you put your name to on your trust's form, are yours and are unaffected by anything here.