Interview training, everything in the room

SIX TOOLS. ONE HABIT.

Every one of these exists because of the same gap: you rehearse in your head, where every answer sounds fine, and the first time you say it out loud is the time that counts. An AI interview tool is only worth having if it hears you say it.

4Dimensions scored8Speaking formats2Interview modes£0To start
1Practise

INTERVIEW RUNS, OUT LOUD.

Paste a role or a job description and the questions are built from what it actually asks for. Answer with your voice, and get scored answer by answer.

Two modes. Practice lets you replay and retake. Live runs start to finish with no pausing, the closest thing to the real room.
Three lengths. Quickfire at five minutes, a ten-minute screening round, or a full 45 to 60 minute interview.
Your CV, not the job title. Attach one and the questions draw on what you have actually done.
Your last run
4.4out of 10↓ 0.7 vs last time
Content4.3
Structure4.0
Relevance4.6
Delivery & pace6.1
Strong instincts, but three answers had no numbers in them.
2Speaking

EIGHT FORMATS, NOT ALL INTERVIEWS.

Nothing here is graded against a right answer. You get coached on pace, timing, emphasis, shape and presence, on whether the point landed.

From 30 seconds to an hour. Elevator pitch through to a full conference talk or a workshop you have to run.
The hard ones too. Difficult conversations, pushback, saying no, rehearsed where the stakes are nothing.
Scored on delivery. Words per minute, filler words, where you sped up and what it cost you.
The eight
Impromptu1 to 5 min
Elevator pitch30s to 2 min
Presentation opening2 to 5 min
Difficult conversation2 to 10 min
Storytelling2 to 10 min
Status update1 to 5 min
Full talk10 min to 1 hr
Teaching session15 min to 1 hr
3Achievements

KEEP THE RECEIPTS, NOT THE MEMORY.

When an answer goes vague under pressure it is almost never nerves, the specific figure was never written down. This is the unglamorous fix.

One number per story. What you did, when, and the result that proves it happened.
Feeds everything else. The same log tailors your CV and supplies the specifics your answers keep missing.
Flags the gaps. Anything logged without a number is marked, because that is where you will fumble.
On record
Rebuilt the launch process across 9 marketsAcme Ltd · Mar 2025£2.4m pipeline9 markets
Ran the pricing change commsAcme Ltd · Jun 2024No numbers yet
4Jobs

RANKED BY FIT, NOT BY KEYWORD.

Live openings from real boards, scored against your keywords and your CV rather than a job title, then practise for the specific one before you apply.

Fit is a ranking aid. Not a decision. Always confirm the details on the employer’s own page.
Practise the posting. One click turns any match into a run built from that job description.
Track a few roles at once. Keep separate searches for the different jobs you are actually going for.
Matches
Director of Product MarketingGentrack · London97%fit
Sr Product MarketingJanus Henderson · London94%fit
Product Marketing ManagerClarivate · London69%fit
5CV checker

YOUR CV, AS A MACHINE READS IT.

A free CV checker that parses your CV the way a recruiter’s applicant tracking system does, and shows you every field it extracted next to every one it could not. It is the file that is checked, not a model’s opinion of it.

Contact details in the body. A phone number inside a running header is skipped by a great many parsers, and the check shows you it sitting there.
Titles, dates and headings. Mixed dashes break an employment timeline. Headings a parser does not know are headings it does not find.
No invented score. No applicant tracking system returns a compatibility percentage, so you get the fields and the reasons instead.
Your last run
4.4out of 10↓ 0.7 vs last time
Content4.3
Structure4.0
Relevance4.6
Delivery & pace6.1
Strong instincts, but three answers had no numbers in them.
6Review

NOT THE SCORE. WHAT MADE IT.

A number on its own tells you where you are and nothing about what to do on Monday. Review joins the two: every habit names the score it drags down, and every score names the habit behind it.

A work list, worst first. The habits behind the score, ranked by what each one costs you, with a drill for every one.
Quoted from your own answers. Evidence is verbatim. A quote that cannot be found in what you actually said is dropped rather than shown to you.
Interviews and speaking, apart. They are scored on different things, so they get their own tabs rather than one average that means neither.
Interview-ready is 7.0
6.4Latest5.1Best4To work on
Answers start before you know where they end. In 4 of 5 answers, and it is what drags structure to 4.2.
7Answering

OUT LOUD, OR TYPED.

Speak your answer and you are scored on all four dimensions. Type it and you are scored on three, because delivery and pace need speech. Both are a full run, and you can switch mid-session.

Say it, or write it. Typed answers are scored on content, structure and relevance. Delivery and pace need something spoken.
Nothing to install. Browser speech recognition, and Chrome or Edge work best. Text mode works anywhere.
Questions on screen. Each one appears when it is your turn, so the first time you read it is the time you answer it.
Live now

“So, tell me about a time you had to change someone’s mind.”

0:04
SpeakingTyping
Answer either way · switch any time

START WITH ONE RUN.

Three runs a month, free, no card. Enough to find out what you actually sound like.