HomeGroundworkMock interview

Timed · scored · follow-ups included

A mock interview that pushes back.

The whole loop, round by round: the recruiter, coding graded by real tests, the deep dives, system design, behaviour and the number. Answer out loud against a clock, get pushed the way real interviewers push, and leave with a debrief that decides like a hiring committee.

12 rounds247 questions with model answers955 coding problems with tests

Four steps, then the debrief

No microphone, nothing recorded: you answer out loud to yourself, against a clock, the way you would in the room.

  1. Plan the loopPick the role, your experience and the kind of company. The rounds, their order and their weight follow real loops.
  2. Answer out loudEach question has a clock. Say the answer as you would in the room, and jot the skeleton if it helps.
  3. Get pushedThe interviewer follows up the way real ones do. That is where a prepared answer runs out.
  4. Read the debriefMark yourself against what the round is listening for. Core rounds can sink the loop, just like a real committee.

What do you want to practise?

Whose loop?

How ready you are

Nothing to measure yet.

Finish one round and this fills in: a readiness number out of 100 that leans on your recent sessions, your streak, how each skill is moving and which round to practise next. It stays in this browser.

Every round, decoded.

Who sits across the table, what they are actually deciding, the shape of an answer that passes and what gets marked down. Read it before you walk in.

Meera · Talent partner

Screening call

10 questions

Twenty minutes with someone who cannot judge your code but can end the process today.

Across the table
Recruiter, non-technical
What they decide
Whether you exist

The shape of an answer that passes

  1. Who you are in 30 seconds
  2. Why this role
  3. Notice period
  4. A range, not a number

Do this

  • Open with a 30-second story: current role, the one thing you are proudest of, and why you are looking.
  • Know your notice period, your current CTC and your expected range before the call. Say the range, not a single number.
  • Ask one real question about the team. It is the only signal they have that you are choosing too.

Marked down for

  • Rambling past two minutes on any answer
  • Naming a number before they have named a band
  • Rambling, or a bad number

The rubric is the same one the room uses.

After each spoken answer you mark yourself against the model answer on five things. Coding rounds are marked by the tests, less a little for every hint you open.

  1. You answered what they were really testing

    Yes only if you named the idea underneath the question, not just its surface.

    20%
  2. Your answer had the substance of the model answer

    Check your main points against the model. Missing its central idea is a No.

    30%
  3. You stayed clear of the answer that loses the room

    If anything you said sits close to the red box, mark No.

    20%
  4. You held up when they pushed

    You answered the follow-up itself, without falling back to your prepared answer.

    20%
  5. Structured and concise — no rambling, no filler

    A clear first sentence, the core in about two minutes, and no circling back.

    10%

From score to verdict

No hire0–44%
Lean no45–64%
Hire65–79%
Strong hire80%+

Then the committee decides the loop

Core rounds count double

Coding and behaviour are always core, plus the rounds your role lives on: the frontend rounds for frontend, backend for backend, design for seniors. They carry twice the weight in the average.

A no-hire in a core round sinks it

However well the other rounds went, a clear no in a core round is rarely overruled in the debrief.

Two lean-nos read as a no

One weak round can be argued away. Two, and the loop is a lean-no unless someone in the room fights for you.

Design decides the level

Pass everything but come in under the bar on design, and the offer lands a level down. Strong-hire every round and there is a case for a level up.

A debrief, not just a score.

Every loop ends the way a real one does: a verdict, the level it would land at, the reasons a committee would give, and exactly what to study next. Weak answers go on a retry list until you nail them.

  • A verdict per round and for the loop
  • The level call, below, at or above what you asked for
  • Chapters to read next, picked from your weakest skill
  • Your trend across every loop you have run
Sample

Full-stack · mid-level · product company

Hire, at mid-level

  • Screening call86%
  • Coding72%
  • JavaScript & TypeScript81%
  • React & the frontend64%
  • System design69%
  • Behavioural77%

Strongest round: Screening call (86%). Weakest: React & the frontend (64%).

Next: re-read Why React re-renders and Effects in depth, then retry the four answers under 60%.

Questions people ask first.

Do I need a microphone?

No. Nothing is recorded. You answer out loud to yourself, as you would in the room, and can jot the skeleton in the notes box. Saying it out loud is the practice.

Who marks my answers?

You do, against the model answer and the five-point rubric, right after you answer. Coding and machine coding rounds are marked by real tests running in your browser.

How long does a loop take?

Pick Quick, Standard or Full. The planner shows every round and the total time before you start, and a single round takes about as long as one real one.

Can I stop half-way?

Yes. The interview saves as you go. Come back later and walk straight back in to the question you left on, or throw it away and start fresh.

What happens to weak answers?

Anything you score under 60% lands on your retry list. Score 80% or better on it later and it comes off.

Is any of this sent anywhere?

No. Your loops, scores and notes stay in this browser. There is no account and nothing to sign up for.