Practise interview answers in writing
Turn the job description you saved with an application into eight questions, write your answers, and see how each one scores against a worked sample.
Five free credits every month, renewing automatically. No card needed to start.
Interview practice usually means reading a list of questions and thinking, yes, I could answer that. Then in the room it comes out as a shrug and a story that goes nowhere. Writing the answer down is slower, and that is the point of it. Career Auntie takes the job description you saved against an application, builds eight questions out of it, and gives each one a worked sample answer. You type your own answer underneath and it is scored against that sample, with written feedback on what yours was missing. Nothing is recorded and nobody is listening. This is practice at a keyboard, on your own, against one specific advert.
Eight questions, drawn from the advert you saved
Career Auntie reads the job description on the application and picks the four competency areas it leans on hardest. Each becomes a topic. Each topic gets two questions: one technical, about the methods, tools or judgement the role actually names, and one behavioural or situational, about an occasion.
Generic questions are ruled out deliberately. "Tell me about yourself", "where do you see yourself in five years" and "what is your greatest weakness" are excluded, because rehearsing an answer to those teaches you nothing about this job.
Every question arrives with a sample answer of 80 to 100 words. Behavioural samples follow the situation, task, action, result shape. Technical samples set out an approach and then ground it in a concrete case. The sample is a model of the shape, not a script to learn: the facts in it are not yours, and an interviewer can tell.
The week between the email and the interview
"They have asked me in on Thursday and I have not interviewed since 2019." The format has not changed much. Your practice at answering it has.
"I know the job. I talk myself out of the answer on the day." Having written a version down means there is something to reach for, even when what you say is not quite it.
"I have three interviews in a fortnight, for three different roles." Each prep sits on its own application with its own advert behind it, so the questions for one do not bleed into the next.
"I always run out of road halfway through an example." That is a structure problem rather than a knowledge one, and structure is the thing the scoring is most obviously looking at.
The application comes first
- Add the application before anything else. Interview prep hangs off an application, not off a page of its own. In the tracker, add the company, the role and the date, then paste the advert into the job description field. That paste is what the questions are built from, so an application saved without it has nothing to read.
- Open that application and start a prep. The button sits with the other actions on it: view the posting, update the status, edit the notes, create the interview prep.
- Read the overview. Four topics, eight questions, and counters for how many you have answered and your average across the ones scored so far.
- Answer one in writing. Type into the box in your own words, at your own length. The answer saves against the question, so you can leave it and come back.
- Score it, if you want it scored. Scoring is a separate step and a separate charge. It compares your answer with the sample for content, for structure, and on a technical question for correctness, then returns a mark out of 100 with written feedback.
- Rewrite, then score again. Editing an answer marks the old score stale, because that score belonged to the previous wording.
One question, answered twice
Example
The advert
A housing association hiring a complaints officer. The advert asks for case ownership, decisions that hold up on appeal, and someone who can say no without escalation.
The question
Behavioural, on unwelcome decisions.
Tell me about a time you gave a customer a decision they did not want, where the policy was clear but they would not accept it. What did you do, and how did it end?
The sample answer supplied
A tenant had been refused a bathroom repair under the planned works policy and had chased it for three weeks. The case came to me and the answer was still no. I rang rather than wrote, took them through the part of the policy that applied, and gave the date the survey was already booked for. The leak they mentioned in passing was covered, so I raised it as a separate job. The complaint closed without going to appeal.
Badly, then well
Badly.
I am good with difficult customers. I stay calm, I listen properly and I try to see it from their side, and usually people come round.
There is no occasion in it. No situation, no task, no action, no ending. Every sentence is a claim about the sort of person you are.
Well.
Last spring a compensation claim reached me after two people had turned it down. The answer was still no. I rang instead of writing, took the customer through the clause that applied and why the meter reading mattered, and did not pretend it was out of my hands. I waived the late fee built up while they waited, on the call.
One occasion. A constraint you could not move. Actions you took, including fixing the thing next to the one you could not fix. The scoring reads for structure and specifics, not confidence, which is why the surer-sounding version does worse.
No mock, no microphone
There is no mock interview here and there is no voice practice. Nothing listens to you, nothing records you, and no simulated interviewer takes a turn. It is written questions, written answers and written feedback. If you want to hear yourself say it out loud, that happens in your kitchen.
The questions are generated from an advert, not leaked from an employer. They are a reasonable reading of what the competencies ask for. The panel will ask other things as well.
The mark is a comparison against one sample answer, produced by AI. No interviewer sees it and it is not a forecast of how the day will go. It reads whether your answer had a structure and some specifics in it, which is not the same as being ready.
0.8 credits a set, 0.25 an answer
Generating a set for one application costs 0.8 credits. That is four topics, eight questions and eight sample answers, in a single run.
Scoring costs 0.25 credits, charged per answer rather than per set, so you can score the two you are worried about and leave the rest. Every account gets five free credits a month, renewing automatically, with no card needed to start, which covers a first set and several answers. Credits come off when you submit, so an abandoned run still spends them.
Common questions
Can I use it without adding an application first?
No. Interview prep hangs off an application, because the questions are generated from the job description saved on it. Add the company, the role and the advert in the applications tracker, then open that application and start the prep from there. Adding an application is free and takes three fields.
Is there a mock interview or voice practice?
No, neither exists. There is no simulated interviewer, no microphone and no recording. Practice here is written: you read the question, type an answer and read the feedback. If you want to rehearse out loud, use the sample answers as a guide and do that part yourself, away from the screen.
What is the score actually measuring?
How closely your written answer matches the sample answer for content, for structure, and on a technical question for correctness. It is a mark out of 100 produced by AI. It is not an employer's view, not a forecast of the interview, and not a measure of you. Treat it as a reading of the answer.
What happens if I rewrite an answer?
The saved score is marked stale, because it belonged to the previous wording. Rewriting costs nothing. Scoring the new version costs 0.25 credits, and the fresh score replaces the old one on that question, so the average you see on the overview follows your latest attempt rather than your first.
What if the advert is short on detail?
Then the questions come out broader, because there is less in the advert to anchor them to. A job description carrying the responsibilities and the essential criteria produces sharper topics than a two-line listing. Paste the fullest version you can find, up to 10,000 characters, while the advert is still up.
Will these be the questions I am asked?
No, and nothing could promise that. They are built from the advert, so they cover the ground the employer said matters, in the shape a competency interview usually takes. The value is in having written an answer about that ground, not in matching the wording of what the panel says on the day.