Product Tour

See exactly how returnzero_ works.

One loop, four moves: learn the concept, practice it in a real sandbox, defend it in a live mock interview, and keep it with spaced repetition. Every tool runs right in your browser.

Explore problem catalog View pricing

Try It

Run every tool yourself

Run code against a test suite, flip a review card, drive a mock interview, and see your plan on the dashboard.

Open the dashboard

credit_tracker.py

Python

1def credit_tracker(events):
2    # Track allocations and client accounts
3    credits = {}
4    for ev in events:
5        if ev["type"] == "allocate":
6            credits[ev["client"]] = credits.get(ev["client"], 0) + ev["amount"]
7        elif ev["type"] == "consume":
8            current = credits.get(ev["client"], 0)
9            if current >= ev["amount"]:
10                credits[ev["client"]] = current - ev["amount"]
11    return credits

Test Results

Run the tests to verify the implementation.

The Loop

Learn, practice, interview, review

Every track runs on the same four steps - learn the concept, build it in code, defend it in a mock interview, and keep it with spaced repetition.

Learn the concepts

Start from deep breakdowns - mobile and distributed system design, AI engineering, infrastructure, and DSA - with the real trade-offs and capacity math worked end to end.

Practice in the sandbox

Solve build-style challenges in an in-browser editor. Write code, run the real test suite, and iterate - Python, JavaScript, and TypeScript, no setup.

Run a mock interview

A live AI interviewer runs a full round: clarify the problem, code your solution, then defend complexity and edge cases - and get a written performance summary.

Lock it in with review

Active-recall cards scheduled by the SM-2 algorithm resurface what you're about to forget, so the concepts stick past interview day.

Know What's Graded

See the interviewer's own scorecard

Every coding challenge ships an interviewer guide - the signals they look for, the green / yellow / red flags, and the 1-5 rubric. Stop guessing what good looks like.

Open a sample guide

Signals the interviewer looks for

Green

Names the stack unprompted and explains the LIFO connection before writing a single line
Re-seeds the sentinel correctly on an unmatched closer

Yellow

Uses three if/elif blocks for bracket matching instead of a dict - correct but verbose
Does not handle the empty string explicitly but arrives at the right answer coincidentally

Red

Attempts to solve Level 1 with a counter (count += 1 / count -= 1) and cannot handle ([)] without rewriting
Cannot explain why the sentinel -1 makes the formula index - stack[-1] work for the first run

Scored 1 - 5 on five dimensions

Approach clarityCorrectnessCode qualityEdge case handlingComplexity analysis

Approach clarity

Named the stack and explained the LIFO connection before writing code. Described the sentinel trick for Level 3 before implementation. No hints needed.

Still using a counter after being told it fails on interleaved brackets; could not reach a correct design.

Excerpt from the Valid Parentheses interviewer guide.

See It For Real

Open an actual problem

Real pages from the catalog, open to read right now - here are three to start with.

Browse all problems

Mobileeasy

Free

Weather App

Design a forecast app like the Google or Apple weather app - one screen for the current location, a forecast that loads instantly from cache, an honest offline state, and a constrained background refresh.

mobilecachingofflinelocation+1