Gains:
- Ability to transform scattered notes after an individual interview into a structured, unbiased and observation-based record with the support of artificial intelligence
- Ability to write requests for meeting summary, process follow-up and next step plan under expert control.
- Understand the risk of AI adding and tagging comments and see that the record must be anonymised and verified, separating fact from interpretation
When an individual interview is over, the expert is often left with a messy note: sentences scribbled in the margin, a few memorable phrases, half-thoughts of what to do next. Turning this note into an organized, objective and traceable record of the conversation takes time and often extends to the end of the day or even the next day; Meanwhile, details are lost. An interview record is a professional document that documents when an interview with a student took place, for what purpose, what was observed, and what steps were planned. Good record kept; It ensures continuity, gives common ground to collaborating colleagues, and enables accountability when necessary.
In this unit, we will learn to use AI as an editing assistant that transforms the messy post-interview memo into a structured, unbiased and observation-based record. But let's put the most critical principle from the beginning: AI organizes the grade; It does not interpret, label or diagnose the student. The facts, observations, and decisions that form the backbone of the record come from the expert.
Anatomy of a good recording: separating fact from interpretation
The most basic professional principle of record keeping is to separate fact from interpretation. Phenomenon; is what is observed or said (“the student avoided eye contact throughout the interview”, “he said his grades dropped on the last exam”). Comment; is the professional meaning that the expert derives from these facts (such as "signs of anxiety observed"). Confusing the two makes the record both biased and misleading. For example, “lazy student” is a label, not a fact; "The student stated that he did not submit his last three assignments" is a fact.
AI is a two-way tool here. The good: can organize a messy note into a neat structure that separates fact and comment headings. Risky aspect: tends to spontaneously add comments and tags; He may sprinkle words like "unhappy", "introvert", "troubled" without you wanting it. That's why you should always limit your prompt to "don't add comments, only edit facts I wrote."
Attention: Do not record any comments produced by the AI unless you professionally agree with that comment. You are responsible for every sentence of the recording; "The AI summarized it like this" is not a justification.
Anonymization: prerequisite for registration
Call records contain sensitive personal data. Be sure to anonymize a grade before giving it to the AI: make the name "Student A", the class "a middle school student", the parent "a parent"; Remove school name, identifiers other than date. You add the actual credentials while saving the regular text the AI produces to your own secure system. So AI only enters anonymously into the “language and form” phase of the process; The authenticated record never goes to a public tool.
Step by step: from messy note to organized recording
- Collect the raw note. As soon as the interview is over, write down what is on your mind; Do not seek order yet.
- Anonymize. Replace credentials with general phrases.
- Separate fact/interpretation. Roughly mark in your notes which is an observation and which is an inference.
- Have it edited by AI. Ask it to fit into a structured template; Ban him from adding comments.
- Check and complete. Read the output; Remove any misinterpretations, add the missing facts, and decide the next step.
- Save to secure system. At this final stage, add the credential to your closed system.
Copiable templates
1) Configuring messy note:
Your role: WRITING assistant to a school counselor.Rule: DO NOT ADD COMMENTS, diagnoses, or tags. Just edit the note below. Put the anonymous raw interview note below under the following headings: 1) Purpose of the interview, 2) Observed facts (what the student said/observed), 3) Expert note (leave BLANK, I will fill in), 4) Planned next step (leave BLANK). Raw note: [paste your anonymous note]
2) Interview summary (for continuity):
Condense the anonymous conversation recording below into a neutral 3-4 sentence summary that my colleague can read in 30 seconds. Using labels and diagnoses; Include only the purpose of the meeting, the topic discussed and the planned steps.Recording: [paste anonymous recording]
3) Process tracking table:
Below is a summary of 3 interviews with the same student (anonymous). Arrange them in a table with columns for date, topic covered, change observed (at the phenomenon level), and planned steps. Add comments.Summaries: [paste anonymous summaries]
4) Open-ended interview question generator:
Your role: PDR writing assistant. I am preparing for a supportive conversation with a 15-year-old student about academic motivation. Suggest 8 non-judgmental, open-ended questions that make it easier for the student to express himself. Avoid closed-ended or leading questions.
Weak prompt / Strong prompt
Weak:
Summarize this meeting: Ali 9-A, his father is unemployed, very aggressive, I think he has an anger problem.
It contains identifying information, mixes fact with interpretation (“very aggressive”, “has anger problems” are labels), and invites the AI to make a diagnosis.
Strong:
Your role: PDR writing assistant, add comments/diagnoses.Edit raw note anonymously. Keep the facts, leave my comment line blank.Raw note: 15 year old student; There were two moments in the meeting when he raised his voice, he said that he had received warnings from the class three times in the last two weeks, and implied that there was tension at home. Purpose of meeting: support after behavioral warnings.
Anonymous, fact-focused, limited and for regulatory purposes.
three mini cases
Case 1 — Capturing the tag. An expert had AI edit notes from an intense week of 12 interviews. In the printout, AI had spontaneously added the phrase “introverted and prone to depression” for one student; However, in the expert's note there was only the fact that "he spoke little in the interview". The expert deleted this tag and preserved the case. This prevented a child's record from being marred by an unfounded diagnosis.
Case 2 — Continuity gain. A guidance counselor was having difficulty remembering each of the five meetings she had with the same student in three months. He had the AI sort the anonymous summaries into a process table; The pattern became clear that student absenteeism decreased but anxiety increased during exam periods. This pattern allowed the specialist to focus the next meeting on anxiety management. The expert made the decision; AI has only made scattered information visible.
Case 3 — Time recovery. A PDR specialist spent an average of 50 minutes a day writing records. Anonymous reduced this time to ~20 minutes per day by having AI edit the raw note and doing the final check himself. He devoted the saved time not to increasing the number of meetings, but to better preparing before each meeting; The recording quality did not decrease because the facts and decisions always came from the expert.
Common mistakes
- Working with credentials. Pasting raw note with name. Solution: always anonymize first.
- Adopting AI's interpretation. Recording added labels like "introvert" without realizing it. Solution: fill in the comment line, check the AI's.
- Fact-interpretation confusion. Writing tags like "lazy" or "aggressive" on the recording. Solution: write only what is observed and said.
- The next step is to have the AI install it. Leave the plan to the vehicle. Solution: the expert decides the step.
- Extreme detail. Recording every sentence of the conversation is wrong in terms of both confidentiality and usefulness. Solution: keep the necessary facts that serve the purpose.
Table: Fact or interpretation?
expression
Type
Is it suitable for recording?
"He avoided making eye contact during the meeting"
Case (observation)
Yes
"An introvert child"
Comment/tag
No (in its raw form)
"He said he didn't turn in his last three assignments."
Fact (statement)
Yes
"Depressed"
Diagnosis (non-jurisdictional)
no
"He stated that he could not sleep before the exam"
Fact (statement)
Yes
"His family is indifferent"
Comment/tag
no
In summary
The interview recording is an objective and traceable professional document that separates fact from interpretation. AI is a powerful helper in structuring, summarizing and process charting the scattered raw note (after anonymizing); but the comments, labels and decisions come from the expert. Anonymously introduce AI into form and language; Keep the identified record in your own secure system.
Application task
Completely anonymize the notes of a recent (or fictional) conversation you took. Have the AI organize it with the "structure messy note" template above. Examine the output: Did the AI add a comment or tag that you did not write? Mark your additions, delete them, and create your own fact-oriented final version.
checklist
- [ ] I anonymized all identifying information before feeding the raw note to the tool.
- [ ] I wrote the "add comment/diagnosis/tag" limit in the prompt.
- [ ] I strip out the comments added by the AI in the output and preserve only the facts.
- [ ] I made the comment and next step decision myself.
- [ ] I wrote the authenticated record only to my own secure system.