In My Words

Start with
their words

A short brief before every session with young people aged 13 to 17, designed so you can check every line of it yourself.

Your workflow

  1. Create a visit

    Add the young person's first name and appointment time, and you'll get a private link to send them.

  2. Read the brief

    It arrives once the young person has reviewed and sent it, and takes a minute or two to read.

  3. Write your plan

    After the visit, write the plan the way you'd normally write your note.

  4. Approve

    Read through the plain-language version, edit anything you like, and send it.

How to read a brief

A few things worth keeping in mind each time you open one.

  1. The quotes are the young person's exact words.

    The phrases above them are the AI's paraphrase. If the two ever seem to disagree, go with the quote. Code checks every quote against the young person's writing before the brief is sent.

  2. "Ask about" marks wording that could mean more than one thing.

    The AI was told not to guess, so treat it as a good opening question for the visit.

  3. A missing topic doesn't mean the young person denied it.

    They chose what to send, and anything they removed was deleted. "Not covered by the questions" is about the questions they answered.

  4. Safety items stay in no matter what.

    A code rule keeps them in the brief and adds them if the AI missed one. It looks for specific phrases and can miss indirect ones, so screen the way you normally would.

  5. There are no scores or diagnoses here.

    How much a young person wrote tells you nothing about how they're doing.

A little AI literacy

Automation bias is well documented. People tend to lean on automated suggestions more than they should, and a systematic review found the same pattern with clinical decision support (Goddard et al., 2012). That's why the brief always shows its evidence. It helps to read the quotes and treat the phrases as a first draft.

Some people share more with a computer. In one study, adults told a virtual interviewer more when they believed it was fully automated than when they thought a person was operating it (Lucas et al., 2014). There's much less evidence for young people, so it's best not to assume a young person will write more than they'd say out loud, or that the brief is the whole story.

The AI can be wrong in quiet ways. It might soften something, overstate it, or leave it out. Word-for-word quotes, flagged ambiguity, the safety rule, and the young person's own review all make that less likely, though none of them can rule it out completely.

What the young person was told

  • You'll read the brief before the visit. Nobody watches it in real time, and if they need help right away, they should call or text 988.
  • Anything about someone hurting them, or about hurting themselves or someone else, stays in the summary, and you may need to act on it.
  • Depending on the situation and the law, a parent, guardian, or another adult whose job is keeping them safe may be involved.

See every screen the young person sees