Skip to content

Writing, editing and languages

AI interviews for writers, editors and translators

Talk through your real writing, editing or translation work with an AI interviewer, then see your results first and choose who sees them. Below: what strong answers show, the 5 practice questions, and a worked example.

  • Free for experts
  • About 15 minutes, whenever you like
  • Type or speak. No camera needed

Retake it up to 3 times: you keep every attempt, and choose which one represents you.

Who it's for

The work it covers

Writers, editors, content strategists, translators and localization specialists.

Example roles

  • Content writer
  • Copywriter
  • Technical writer
  • Editor
  • Content strategist
  • Translator
  • Localization specialist
  • Journalist

Name your exact specialty when you apply, and the questions follow it.

Topics the interviewer draws on in writing, editing and languages

  • A piece they are proud of and what made it work
  • A major edit or rewrite and why it was needed
  • Writing for readers who knew less than they did
  • Researching an unfamiliar subject and checking facts or terms with sources or experts
  • A factual error or mistranslation and how it was caught
  • Adapting tone and style for a brand, audience or culture
  • Editor or client feedback they disagreed with
  • Working to a tight deadline without losing accuracy

Scoring

What a strong answer shows

These are the words the scoring AI is given for writing, editing and languages, as written in Aethel's configuration. The scoring AI is told to score an area below 60 when its answers show one of the warning signs, however confident the answer sounds.

Strong work

The candidate checks facts and terminology against primary sources, explains how each edit serves the reader, adapts tone and structure on purpose, and can show the work was accurate and understood.

Warning signs

The candidate does no research or fact-checking, rewrites until a client is satisfied without knowing what was wrong, or cannot explain any choice they made.

The four scored areas

  • Editorial judgment

    Problems found, edits made

    Counts as evidence when the candidate describes a specific writing, editing or translation problem: what was wrong, the options, the choice they made, or how they knew it worked.

  • Craft and accuracy

    Structure, sourcing, tone, terminology

    Counts as evidence when the candidate describes how they produce accurate, well-structured work: research, sourcing, fact-checking, structure, style guides, tone, terminology or translation choices.

  • Communication

    Briefs, edit notes, pitches

    Counts as evidence when the candidate describes explaining their work or decisions to other people: briefs, edit notes, pitches, style guidance, or walking a client through changes. Answers that respond to the question asked, in a clear order, with a concrete example, also count, but only up to 80 on their own. Grammar, word choice and fluency are never judged.

  • Collaboration

    Editors, clients, subject experts

    Counts as evidence when the candidate describes working with editors, authors, clients, subject experts, designers or reviewers, including handling feedback and disagreement. Saying "we" is not, by itself, collaboration evidence.

An area is scored only when the AI can quote you; otherwise it shows Not enough evidence, not a low score. How AI interviews are scored

Practice

The 5 practice questions

Questions like these come up in the interview. The free practice room asks 4 of them, at any level. The real interview follows up on your own answers instead.

  1. Question 1

    Tell me about a piece of writing, editing or translation you are proud of, and what made it hard.

  2. Question 2

    What did you cut or change, and why?

  3. Question 3

    Imagine a client sends back your draft saying it doesn't sound like them, with no other notes. What do you do first?

  4. Question 4

    How would you explain a major edit to an author who is attached to their original wording?

  5. Question 5

    Tell me about one piece you checked for accuracy before it went out. What did you check, and what did you catch?

Answer 4 of them in the practice room

No account. Each answer gets a quick rule check in your browser.

Worked example

One answer, read by the rules

Made up by Aethel to show how the published rules read one answer: not a real candidate, and not scored by the AI. A real score can differ.

The practice question

Tell me about one piece you checked for accuracy before it went out. What did you check, and what did you catch?

The made-up answer

The clearest example is a 40-page benefits guide I translated from English into Spanish for a hospital group with about 3,000 staff, a job I did through a translation agency. I check in three passes. First, everything that can be checked mechanically: every number, date, dollar amount and phone number against the source, with a quality-assurance tool and then by eye, because the tool misses a figure that is written out in words. Second, terminology, against the client's glossary and the insurer's own Spanish documents. On that guide I found the source used 'deductible' and 'copay' interchangeably in two sections, and I queried the client instead of picking one, because choosing wrong would have told staff the wrong amount they owe. Third, I read the Spanish on its own, without the English, out loud, to catch anything that is accurate but reads like a translation. The client's review came back with two corrections, both style preferences, where their earlier guides had averaged about a dozen. The agency then made my three-pass check standard for all its benefits work.

Marked words are the quotes the rules would look at.

How the rules read it

  1. The situationLevel 2 of 2

    A specific case with where it was and how big: one 40-page guide for about 3,000 readers.

  2. The reasoningLevel 2 of 3

    One alternative weighed against another, with a risk guarded against: querying the client against choosing a term, because a wrong choice misstates what staff pay.

  3. The resultLevel 3 of 3

    Measured, and then adopted beyond the case: corrections counted against earlier guides, and the check became the standard for other projects.

Together the three ratings give 7 of 8 points, which is 91 on the fixed scale. Its longest complete quote has 33 words, so the quote rule does not lower it. It is evidence mainly for Craft and accuracy.

What would raise it

Comparing a second option for the terminology problem, such as following the insurer's documents without asking, and saying why it was rejected would raise the reasoning to 3 and the area to 8 points.

The same story, told vaguely

I always check my translations carefully before they go out. I compare the numbers and dates against the source, check terminology against the client's glossary, ask questions when something in the source is unclear, and read the final version on its own. Clients are usually happy with the result and have very few corrections.

No specific case: this is general practice, so situation is 0 and the area is at most 60. 'Very few corrections' is a vague result.

A whole result: the sample result. More examples: strong interview answers.

Company interview links

Roles companies can set up

Starter templates a company can use for an interview link, and what each one steers the interviewer toward.

  • Content writer

    Focus: Articles and web pages, research, editing, meeting deadlines

  • Editor

    Focus: Structural and line editing, fact-checking, style guides, feedback writers can use

  • Grant writer

    Focus: Researching funders, writing proposals, budget narratives, reporting to funders, working with program staff

  • Technical writer

    Focus: Product documentation, working with engineers, writing for different readers

  • Translator

    Focus: Accurate translation, terminology and style guides, adapting tone for the audience, review and quality checks

FAQ

Questions people ask

How do I show editorial judgment when the interview cannot see my work?

Describe one piece and one decision: what was wrong with the draft, the options you had, what you chose and how you knew it worked for the reader. A short before-and-after line can help. Editorial judgment is scored on the problem, the choice and the result, not on how polished your spoken answer sounds.

What counts as a result for writing, editing or translation work?

Something measured or confirmed: fewer corrections at review, an editor or client accepting the change, readers finishing a task, a support question that stopped coming in, or a correction rate over several projects. The result rating counts a result confirmed by someone else, so an editor's sign-off counts. 'The client loved it' is vague.

Can I talk about editor or client feedback I disagreed with?

Yes. Collaboration covers working with editors, authors, clients and subject experts, including disagreement. Say what the note asked for, why you disagreed, what you proposed instead and how it was settled. Rewriting until a client is satisfied without knowing what was wrong is one of the field's warning signs, so show how you found out what the reader needed.

Can I share writing samples?

Not in the interview. You can add a link to your samples to your profile when you apply. The interview is a conversation in English or Spanish, so it does not test your writing itself or how well you translate. Your scores come only from what you say.

Do I have to share confidential details?

No. The interviewer is told not to ask for names or confidential details about clients, patients, students or employers, and to let you describe situations without identifying anyone. Describe client work without sharing anything you agreed to keep private.

How to prepare for an AI interview

What this interview cannot do

It is one conversation of about 15 minutes, so it covers only part of what you can do, and we have not tested whether results predict job performance. The AI can misread an answer, which is why every score shows its quotes and people make the decisions. It is not a certification or license check. No independent bias audit has been done. Writing, editing and languages was not one of the 8 fields in our own test (not an audit) of whether non-native English grammar changes scores.

All known limitsYour rights as a candidate