Your choice also saves this guide’s current worked-example checkmarks. Only this browser; no account or sync. View saved guides
Saving needs JavaScript. Reading and downloading this guide does not save it.
Make a start
One step at a time.
5 steps
Start with the synthetic worked example or other non-sensitive inputs. Run one prompt at a time in your chosen AI tool. Replace any bracketed placeholders with the requested material, check each reply, and supply the source or earlier outputs that step needs. Paste them again if the tool lacks that context.
Set up the worked example
Open
Open a new conversation in a text AI assistant and a notes app. This is a fictional practice exercise, not a real interview or candidate assessment. A personal timer is optional if you later choose to say an answer aloud; no recording or voice feature is required.
Add the sample
In step 1, replace [paste fixture] with all of ‘The starting point’ (Synthetic input), from ‘SYNTHETIC PRACTICE BRIEF.’ through ‘no recording or timing has happened.’ Include the role excerpt, E1–E5 and the practice request. Send steps 1–4 one at a time in the same chat, checking each reply. Step 5 uses the same notes and outlines, then should ask just one practice question and wait for your answer. ‘The intended result’ (Illustrative output) is a model for comparison, not evidence about your own experience or a completed interview.
Start with the synthetic example, then adapt each prompt to your task.
01
Build the evidence ledger
Use truthful notes and the role's stated tasks without filling gaps.
Your starting prompt
Extract the role requirements and map each experience note to supported skills, personal actions, recorded outcomes and unknowns. Mark proposed dates, agreed trials and unrecorded results precisely. Do not infer personality, seniority, qualifications or performance metrics.
ROLE AND NOTES:
[paste fixture]
Expect
A ledger mapping the role's tasks to E1–E5, with personal actions, recorded outcomes and unknowns separated. The role concerns accuracy, handoffs, early risk communication and learning tools.
Check
E1 has 18 entries minus two duplicates, leaving 16 distinct attendees. E2 records agreement to a trial, not its success. E3 records a proposed Friday deadline, not acceptance or completion. E4 does not establish mastery; E5's wrong quantity was corrected before sharing.
If it differs
If the reply fills in a missing result or turns a team agreement into Taylor's leadership, quote the relevant E note and ask for a corrected ledger. Retain unrecorded outcomes instead of inventing a stronger story.
02
Draft behavioral questions
Practice job-relevant explanations, not answers about protected or private traits.
Your starting prompt
Create six behavioral questions covering accuracy, disagreement, deadline risk, learning a tool, catching a mistake and a clear handoff. Match each to the supplied evidence. If one story serves two questions, say so. Do not invent a new experience to create variety.
Expect
Exactly six behavioral questions: accuracy, disagreement, deadline risk, learning a tool, catching a mistake and a clear handoff. Each should point to relevant source evidence.
Check
E1 can support both accuracy and handoff; that reuse should be explicit. There are five experience notes, so six questions do not require six invented experiences. Questions should concern work actions, not private traits or suitability judgments.
If it differs
If a question uses a new story, ask it to match the question to E1–E5 or mark the evidence gap. If an area is missing, identify it and ask for the six requested areas without adding fictional achievements.
03
Write honest answer outlines
Use situation, task, action and result as a structure rather than a reason to invent results.
Your starting prompt
Write one outline per question and a complete first-person sample outline for the accuracy story. Preserve who did what and all unknown outcomes. Include a reflection prompt rather than an invented lesson. Do not turn a proposed deadline into an accepted one or a trial into measured success.
Expect
An outline for every question, plus a complete first-person sample outline for accuracy. Situation, task, action and result organize the story. A reflection remains a prompt where the notes do not supply an answer.
Check
The complete sample is Taylor's fictional practice voice, not your history. Its result is the corrected list of 16 distinct attendees, with no measured attendance or cost effect. A proposed date, agreed trial and completed result must stay distinct throughout all six outlines.
If it differs
If a result is embellished, point to its E note and ask for a narrower outline with the unknown stated. If the AI supplies a personal reflection, ask it to replace that with a question rather than claiming Taylor learned something unrecorded.
04
Practice useful follow-ups
A follow-up can uncover a missing detail without presuming it exists.
Your starting prompt
Add one follow-up per question that tests understanding or evidence. Explicitly allow 'I do not recall' and invite only true additional detail. Create a self-review rubric using clear / needs work / not evidenced with evidence quotes. Do not score or rank candidates or infer personality or hiring suitability.
Expect
One follow-up per question and a self-review rubric using ‘clear’, ‘needs work’ or ‘not evidenced’, with quoted support. The rubric reviews how well an answer explains evidence.
Check
A follow-up should allow ‘I do not recall’. A question about whether the welcome desk used the corrected list cannot assume acknowledgment occurred. There should be no hiring score, personality judgment, accent assessment or candidate ranking.
If it differs
If a follow-up presumes an unknown outcome, ask for an open question. If a rating predicts employability, ask for evidence-based feedback on the answer only, using the supplied non-numerical labels.
05
Audit then rehearse
Read aloud and revise one concrete issue at a time.
Your starting prompt
Audit every outline against the source notes. Flag unsupported numbers, seniority, completion claims, acknowledgments and business outcomes. Return corrected outlines. Then present one question at a time for self-practice; wait for the learner's answer before feedback. A 60–90 second target is optional and not a measured duration.
Expect
Corrected outlines after an evidence audit, followed by one question and a pause. For this sample, type an answer using only Taylor's notes and label it fictional practice. You may first read it aloud to yourself if you wish.
Check
Only you can measure your spoken duration; 60–90 seconds is an optional target, not an observed result. Feedback should quote your answer and check it against the notes. The AI should wait for your reply rather than inventing a learner answer or declaring an interview completed.
If it differs
If the AI answers its own question or judges an answer you did not give, send: ‘Ask only the first practice question and wait for my answer. Do not invent my reply.’ Correct unsupported facts before continuing, and stop whenever you have finished practising.
Optional: improve the result
Run an evidence-constrained mock interview
Ask one question at a time and distinguish new truthful detail from an unsupported embellishment.
Advanced prompt
Act as a practice interviewer using the supplied role and evidence notes. Ask one behavioral question and wait. After each answer, quote supported claims, ask about one unclear detail and identify any conflict with the notes. New details are learner-supplied and require their confirmation; do not assume them false or verified. Offer one revision using only confirmed facts. Give no hiring score, personality inference, candidate ranking or guarantee of success. End with a short list of answers to practice again.
Optional: use your own input+
Practise explaining your own real experience
Start a new text chat with non-sensitive, truthful experience notes and a role description you may use. Omit personal contacts, confidential employer details and names that are unnecessary. Give each story a label. Fill the placeholders below. Check the evidence ledger, then send these messages one at a time: “Create my requested number of behavioral questions tied to the role and my evidence. Cover the agreed practice areas. Say when a story is reused or no supporting story exists.” “Write one honest outline per question and one complete sample outline from a supported story. Separate situation, task, personal action and recorded result; leave missing reflections as prompts.” “Add one open follow-up per question and a self-review rubric using clear / needs work / not evidenced with quotes. Allow me not to remember; give no hiring or personality scores.” “Audit the outlines against my original notes, correct unsupported claims, then ask me just one practice question and wait for my answer.” Answer in your own words. When feedback asks about a new detail, confirm it only if true; the AI cannot verify it for you. Say ‘Next question’ when ready or ‘Stop practice’ to finish. Measure your own speaking time only if you want a timed rehearsal. The original advanced extension stays sample-scoped until it has your role and evidence; it remains self-practice, not a genuine interview or hiring assessment.
Your own-input prompt
Help me build an evidence-based interview self-practice pack. First extract this role's stated tasks and map my labeled experience notes to supported skills, personal actions, recorded outcomes and unknowns. Preserve the actual numbers, sequence, dates and limits; do not turn a proposed deadline or agreed trial into a completed result. Do not draft the full pack yet. We will create the requested questions, outlines, follow-ups and a non-scored review rubric, then practise one question at a time. If evidence is missing, ask for a true example or leave a gap. New details I give later are learner-supplied until I confirm them; do not automatically treat them as verified or false. Review my answer quality only, with no candidate ranking, personality inference, suitability prediction or guarantee of success. Do not contact anyone, submit applications or claim to conduct a real interview.
ROLE DESCRIPTION
[Paste the relevant role requirements without unnecessary identifying details.]
LABELED EXPERIENCE NOTES
[Paste truthful, non-sensitive stories labeled E1, E2 and so on, or use your existing labels. For each, state the situation, your actions, others' actions, recorded outcome and what you do not know.]
PRACTICE REQUEST
[State the question count, areas you want to practise and any optional spoken-time target. Ask the AI to suggest role-relevant areas if you are unsure.]
Check your own result
Every question and outline uses your stories and the role's actual tasks. Your requested count and practice areas replace the sample's six-question selection.
Your figures, dates, contributors and outcomes remain exact; proposals, agreements, completed work and unknowns do not get merged.
Follow-ups invite true detail without presuming it. You can say you do not recall, and reflection is not invented for you.
Practice waits for your answers. Feedback concerns answer evidence and clarity, with no hiring score or profiling.
Only actual spoken practice can establish timing; written drafts and AI feedback do not prove an interview happened or predict its outcome.
Show, then try
See the shape of the result.
This input is synthetic. The output is a written illustration of the intended result, not a verified AI-tool run.
01 / The starting point
Synthetic input
SYNTHETIC PRACTICE BRIEF. Fictional learner Taylor is preparing for an operations-coordinator interview. No real candidate or employer.
Role excerpt: keep task records accurate, coordinate handoffs, communicate risks early and learn new tools. No industry certification or management requirement is stated.
EXPERIENCE NOTES
E1: At a volunteer repair event, the sign-up sheet listed 18 attendees. Taylor compared it with check-in notes and found two duplicate entries, confirmed the duplication with the event lead, and updated the working list to 16 distinct attendees. Taylor told the welcome-desk volunteer which list to use. No effect on attendance or costs was measured.
E2: For a shared equipment cupboard, Taylor proposed sorting labels by item type. Another volunteer preferred sorting by storage shelf. Taylor asked each person how they looked for items; they agreed to a trial using shelf labels plus an item index. No later trial results are recorded.
E3: Taylor said they would finish a supply list by Thursday, then realized on Wednesday that a supplier sheet was missing. Taylor told the lead Wednesday, asked for the missing sheet and proposed Friday as a revised date. Whether the lead accepted Friday or the list was completed is not recorded.
E4: Taylor had not used the team's spreadsheet tool. They followed its internal starter guide on a copied sample, asked a teammate to check one formula and documented the steps. No qualification, mastery level or measured productivity gain is recorded.
E5: Taylor once copied the wrong quantity into a draft checklist, noticed it during a manual comparison, corrected it before sharing and added a second-pass check. No one else had used the draft.
Practice request: six behavioral questions tied to the role, one honest answer outline per question, a complete sample outline for accuracy, follow-up prompts and a self-review rubric. Target spoken length may be 60–90 seconds but no recording or timing has happened.
Copies text only. For image or file tasks, use the source files named in the steps too.
INTERVIEW PRACTICE PACK · Synthetic learner · Self-practice only
1. Tell me about a time you checked a record that looked wrong.
Evidence: E1.
Outline: volunteer event / accurate working list needed / compare sign-up and check-in notes, confirm duplicates with lead, update list and hand it to welcome desk / 18 entries became 16 distinct attendees; no measured business impact.
Follow-up: How did you establish they were duplicates rather than different attendees?
2. Tell me about a disagreement over how to organize work.
Evidence: E2.
Outline: equipment cupboard labels / two different ways to search / ask both volunteers how they looked for items, agree shelf labels plus index trial / agreement to a trial, with later outcome unknown.
Follow-up: What would you check to decide whether the trial worked?
3. Describe a time a deadline was at risk.
Evidence: E3.
Outline: supply list promised Thursday / missing supplier sheet found Wednesday / notify lead, ask for sheet, propose Friday / revised date was proposed, not confirmed; completion is not recorded.
Follow-up: What did the lead decide? If you do not remember, say so.
4. Tell me how you learned an unfamiliar tool.
Evidence: E4.
Outline: new team spreadsheet tool / need to learn its workflow / follow internal guide on a copied sample, ask teammate to check a formula, document steps / steps documented; no claim of mastery or quantified gain.
Follow-up: Which specific feature did you learn? Add it only if you can truthfully supply it.
5. Tell me about a mistake you caught.
Evidence: E5.
Outline: wrong quantity in draft checklist / catch errors before sharing / manual comparison, correction, second-pass check / error corrected before anyone used the draft.
Follow-up: How would you check that the second pass is working without claiming unmeasured results?
6. Tell me about a handoff you made clear.
Evidence: E1, used from a different angle.
Outline: corrected event list / welcome desk needed the right version / tell volunteer which working list to use after lead confirmation / handoff instruction was given; acknowledgment or later use is not recorded.
Follow-up: How did you confirm they were using the right version? Do not invent an acknowledgment.
COMPLETE HONEST SAMPLE OUTLINE FOR QUESTION 1
Situation: At a volunteer repair event, our sign-up sheet contained 18 entries.
Task: I was checking the working attendee list against check-in notes.
Action: I compared the two sources, found two possible duplicate entries, confirmed them with the event lead, updated the working list and told the welcome-desk volunteer which list to use.
Result: The corrected list showed 16 distinct attendees. I do not have evidence of an effect on attendance or costs.
Reflection to prepare: What would I do the same way or differently next time? Add a genuine answer; the notes do not supply one.
SELF-REVIEW RUBRIC (mark clear / needs work / not evidenced, with a quote)
Evidence fidelity: Every factual claim is supported by a note or an additional true detail you provide.
Personal action: Distinguish what you personally did from what the lead or team decided.
Structure: Give enough situation/task context, then actions and the actual recorded outcome.
Outcome honesty: Separate a proposal, a trial agreement and a completed result. State unknowns honestly.
Relevance: Explain the connection to accuracy, handoffs, risk communication or learning.
Clarity: Use understandable language and check spoken length yourself; 60–90 seconds is a practice target, not a measured result.
Use feedback to revise the answer, not to score employability, personality, accent, confidence or suitability.
PRACTICE ORDER
Choose one question, draft from its outline, say it aloud, answer one follow-up and revise one supported detail. E1 supports two questions, but consider finding another true handoff example for variety. Nothing here predicts interview success.
These checks use the worked example and its particular facts. With your own material, check your actual names, numbers, sources and rules; do not reuse the sample’s values as the answer.
Unsaved checks stay in this tab and reset when you reload. If you choose Save guide, its current example checkmarks are kept in this browser when storage is available.
!
A few things to watch for.
Filling the result section with a plausible outcome that was never recorded.
Claiming team actions as your own leadership.
Treating a practice rubric as a prediction of employability.