Open sourceBuilt in public — come contribute

The fastest way to crack your next interview

Practice questions with real solutions, learn system design from structured tracks, rehearse with an AI interviewer, and keep a streak that actually holds you to it.

Free and open source · No card, no trial, no upsell

crackitt.app/practice/coding

Coding Practice

10 questions · 3 solved this week

DebounceMedium
Deep clone an objectMedium
Event emitterHard
Virtualised listHard

Built by engineers who interviewed at

GoogleStripeAtlassianShopifyRazorpayZomatoUber
Why crackitt

Order, not volume

Most prep sites hand you a thousand problems and no order. CrackItt gives you a track, a solution you can actually learn from, and a habit that survives past week one.

Search 200+ questions
Frontend
Backend
System Design
DSA
Start session

A roadmap, not a random list

Structured tracks take you from fundamentals to the questions that actually get asked, one lesson at a time.

Every stack you will be grilled on

Frontend internals, backend systems, databases, caching, and the infra questions that decide the senior loop.

React
TypeScript
Node.js
Postgres
Redis
Kafka
Docker
GraphQL
Next.js
Python
Go
AWS

200+

questions, lessons, and tutorials across every track — each with a solution, not just an answer key.

14 day streak
M
T
W
T
F
S
S

Check in daily. Consistency beats cramming the night before.

How it works

Three steps to an offer letter

No dashboards to configure, no plan to pick. Start on step one today.

01

Pick your track

Frontend, backend, or full stack. Every track is a roadmap, not a random list — you always know what comes next.

Search 200+ questions
Frontend
Backend
System Design
DSA
Start session
02

Practice with real solutions

Machine coding problems and interview questions, each with a worked solution, complexity analysis, and the trade-offs an interviewer will push on.

Medium

Implement a debounce function

function debounce(fn, ms) {
let timer;
return (...args) => {
clearTimeout(timer);
}
}
Solution + complexity explained
03

Rehearse with the AI interviewer

It asks follow-ups based on what you actually said, then tells you where your answer was thin. Fail here instead of on the call.

AI Interviewerlive
How would you scale reads on a write-heavy Postgres table?
Read replicas first, then a cache layer keyed on...
Good. What breaks when replication lags?
Community

You are not prepping alone

Read write-ups from people who just sat the loop, ask the question you are embarrassed to google, and ship a project the reviewer will actually open.

Blog

What I got wrong in my Stripe onsite

12 min read · 148 upvotes

Question

How deep do they go on event loop internals?

23 answers

Project

Rate limiter with sliding window logs

★ 4.8 · 31 reviews

Tutorial

Designing a URL shortener, end to end

38 min video

What people
are saying

“I did fourteen AI mock interviews in two weeks. By the real one, the follow-up questions were boring — I had already been asked all of them.”

PR

Priya Raghavan

Frontend Engineer

Open source

CrackItt is built in the open. Come build it with us.

Every question, solution, and lesson on this site came from someone who had just sat the interview. The best thing you can do for the next candidate is write down what you learned in yours.

Next.js 16React 19Tailwind 4Express 5PostgresKysely

Frequently asked questions

Everything about the tracks, the AI interviewer, and how the prep actually works.

Yes, all of it. CrackItt is open source and free — the full question bank, both learning tracks, the AI mock interviewer, tutorials, community, and streaks. There is no paid tier and no trial to run out.

Frontend, backend, and full stack engineering interviews — coding rounds, machine coding, system design, and the behavioural questions that get asked alongside them.

You pick a track and difficulty, and the interviewer asks follow-up questions based on your actual answers instead of reading a script. At the end you get feedback on structure, depth, and the things you skipped.

Every question ships with a worked solution, complexity analysis, and the trade-offs an interviewer will probe. They are reviewed by engineers who have run these loops.

One question a day beats a panicked weekend. Checking in keeps your streak alive and your dashboard shows exactly what to review next.

Please do. Publish blogs, ask questions, and share projects in the community — good write-ups get surfaced on the dashboard feed. You can also send a pull request: new questions, solutions, bug fixes, and UI work are all welcome on GitHub.

Still stuck? Email us

Practice. Learn. Get hired.

Start with one question today. Your future self, sitting in that interview, will be glad you did.

Get started for free