Loading…
An AI interviewer that pushes back, adapts to you, and tells you if you’d pass.
Free plan, no credit card required
Before you write anything — what’s the brute force here, and what does it cost you?
Nested loops, so O(n²) time, O(1) space. I want to trade space for time with a hash map of value to index.
Good. What happens when the same value shows up twice?
1def two_sum(nums, target):2 seen = {}3 for i, n in enumerate(nums):4 need = target - n5 if need in seen:6 return [seen[need], i]7 seen[n] = i8 return []✓two_sum([2, 7, 11, 15], 9)→ [0, 1]
✓two_sum([3, 2, 4], 6)→ [1, 2]
✓two_sum([3, 3], 6)→ [0, 1]
3 passed · 5 total · 2 hidden
Calibrated to the bar at
You said “we” when the interviewer needed “I”. Your impact was “users were happy” instead of a number. Your conflict story ended with you being right all along. Nobody in the debrief ever tells you that. You just get a rejection email six days later.
It does not accept your first answer. If a story is vague, it asks what exactly you did, what would have happened if you hadn't, and how you measured it. These are the same follow-ups a bar raiser uses.
An Amazon SDE III round scores your answers against the Leadership Principles at a senior bar. A Google L4 round probes Googliness and structured problem solving. The questions and the bar both change.
You get the notes a real interviewer writes after you leave the room: the signal they'd report, what was missing from each answer, and a rewritten version of your weakest story.
Run a single round, or the full loop back to back under one configuration, with the same company and level carried through every round.
A live editor the interviewer reads as you type, running against real tests.
An architecture canvas the interviewer can see, name components on, and push back against.
Questions across competencies, answered by voice, with real follow-up probing.
Walk me through your approach before you write anything.
Live · calibrated to Amazon SDE II
1 def is_valid(s):
2 stack = []
3 for c in s:
4 if c in "([{":
5 stack.append(c)✓ is_valid("()[]{}") → True
✕ is_valid("(]") → expected False, got True
Auto-saved just now
Tracking
Flagged
Missed edge case — empty input string
Watching your last edit…
Choose the company and the level you're actually interviewing for: Amazon SDE II, Google L5, Meta E4. The interview is built from that.
Tap the mic and answer out loud, the way you will on the day. The interviewer listens, follows up, and pushes back when you're vague.
Get a structured debrief in under a minute: hire / borderline / no-hire, per-answer breakdown, your top issues, and stronger rewrites.
A generic tool asks “tell me about a challenge.” LoopReady interviews you the way your target actually does, and the level you pick changes the bar, not just the wording. Pick a company to see the difference — interviewing somewhere else? Pick Custom.
Follow-ups your interviewer will actually ask
Real output from a session, not a mockup. Every point is tied to something the candidate actually said.
Welcome back, Aryan
Run a mock, then read the debrief like a real interviewer would write it.
Interviews
0
11 completed
By round
Latest signal
Plan
Recent sessions
Borderline for Amazon SDE III/L6, with stronger-than-average behavioral structure but not enough senior-level scope. The candidate gave clean STAR answers, accepted personal responsibility, and improved materially when probed for data. The main ding is level calibration.
What an interviewer wanted and did not hear
A hire-level answer needed the manager’s real argument for MongoDB, the evidence you brought, and what you conceded in your own position. The story made you look certain rather than data-driven.
What you said
“My manager wanted MongoDB but I thought Postgres was better. We talked about it and eventually went with Postgres and it worked out fine.”
Stronger version
“He proposed MongoDB because he expected event payloads to change often. I built a one-day prototype on 5M events: Mongo won on ingest, but our three reporting queries ran 2–3× slower. I changed my own position to Postgres with JSONB, and he accepted. p95 stayed under 400 ms all quarter.”
Because reading your own polished summary back to a model is not the skill being tested.
| LoopReady | Chatbot prompt | Friend mock | |
|---|---|---|---|
| You answer out loud | ✓ | — | ✓ |
| Follows up on vague answers | ✓ | rarely | depends |
| Knows your company's bar | ✓ | — | unlikely |
| Scales to your target level | ✓ | — | — |
| Honest verdict | ✓ | flattering | polite |
| Available at 11pm | ✓ | ✓ | — |
One failed loop costs you a year of compounding.
Practice is the cheapest part of the process.
Free
For trying it out
$29$19/mo
For the full loop
$89$69/mo
For the real thing
Video rounds use one credit each. Voice stays unlimited.
| Free | Voice · $19/mo | Premium · $69/mo | |
|---|---|---|---|
| Session frequency | 1 every 7 days | Unlimited | Unlimited |
| Coding round (live editor + execution) | |||
| System design round (canvas) | |||
| Full loop, rounds back to back | |||
| Real-time voice interviewer | |||
| Written debrief | |||
| Session history | Last session | ||
| Video-avatar interviews | 3 / month |
What am I actually paying for?
Access and volume. Every mock runs a real language model for the interviewer and a second pass for the debrief, which costs money each time. Paying removes the daily cap and unlocks the coding and system design rounds, so you can practise the whole loop as often as you need without watching a meter.
Can I cancel anytime?
Yes. Plans are month to month with no contract, and you keep access until the end of the period you've paid for.
What does the Free plan actually get me?
1 mock every 7 days, any round you like -- behavioral, coding, or system design -- in the same real-time voice the paid plans use, with the full written debrief. It is enough to judge whether the feedback is worth paying for. That is the point of it.
What is the difference between Free and Voice?
The interviewer is identical: the same voice, the same probing, the same debrief, and on Free you can already pick any round type. What Voice removes is the 7-day wait between mocks -- unlimited interviews, and full loops with rounds run back to back, which Free's one-at-a-time cap does not allow.
How do video interviews work?
Premium includes video rounds with an on-screen interviewer who sees your code and your diagram and talks with you in real time. Each video round uses one credit, and voice rounds stay unlimited, so you can save video for the practice that matters most.
Do you offer student or bulk pricing?
Not yet, but if you're organising practice for a bootcamp or a university group, get in touch and we'll work something out.
Question we have not answered? support@loopready.io
Yes, and that's the point. Reading a written answer is a different skill from telling the story under pressure. You tap a mic button, answer, and the interviewer replies in voice.