Practice that grades like production
Katabench started with a simple frustration: every practice platform stops at "correct". You submit an O(n²) solution, the checkmark turns green, and you walk away having learned the wrong lesson. Meanwhile the skills that actually distinguish senior .NET engineers — performance instincts, query discipline, architectural judgment — go untrained because nobody grades them.
So we built the platform that does. Your C# is compiled and run server-side in an isolated sandbox, and measured: wall-clock time against a budget, bytes allocated, the query plan the database chose, structural metrics from your source, dependency rules, attack payloads survived. A naive solution doesn't pass here — and the moment it fails, you can see exactly why and exactly how far off you are.
What we believe
Correct is the starting line, not the finish
Production doesn't grade on green checkmarks. It grades on latency, allocations, query counts, and whether the next engineer can change your code without breaking it. Practice should grade the same way.
Feedback beats lectures
A timeout on a hidden 200,000-element test teaches Big-O viscerally in a way no article can. We engineer the feedback loop and stay out of the way.
Real tools, real stack
A real compiler, a real database, real architecture rules. If the exercise is fake, the skill is too — so nothing here is mocked.
Senior skills, not interview tricks
Algorithms matter, but so do killing N+1 queries, refactoring a mess without breaking it, keeping dependencies pointing the right way, and writing code that shrugs off hostile input. We train the whole craft, because that's what the job actually is.
Who's behind it
Katabench is built by Milan Jovanović, a .NET engineer and educator. Questions, feedback, or just want to argue about allocations? Write to [email protected] — every message gets read.
Come get graded 🥋
The Free plan is free forever. The only thing on the line is your O(n²) habit.
Start solving — it's free