Sorry, we made you grind LeetCode

I’m Vivek, co-founder/ceo of HackerRank. We’re sorry we made you grind leetcode. That line is on billboards round SF now.

HackerRank didn’t invent algorithmic interviews, however we helped normalize a model of technical hiring that was months of grinding questions with little connection to the precise job.

Technical interviews have come a great distance: from “How many golf balls slot in a 747?” to studying code over the telephone, writing it on a whiteboard, and finally fixing LeetCode-style issues.

Each step received us slightly nearer to the work. But we had been nonetheless largely evaluating how properly somebody had ready for an interview, not how properly they may construct software program.

AI has modified that work.

Developers more and more plan with brokers, construct with them, and assessment their output. They have to know when the code is nice, when it wants one other move, and when it’s utterly unsuitable.

The hiring course of must evolve too.

At HackerRank, we’re altering three issues:

• Tasks: Algorithmic questions → actual work in code repositories
• Evaluation: Code correctness → AI fluency and judgment
• Candidate expertise: IDE → agentic improvement atmosphere

We’ve packaged this into Chakra, our agentic interviewer.

Candidates get a practical drawback and entry to AI instruments. They remedy it as they might in the true planet. Chakra then digs deeper: Why did you select that method? What would you modify? How did you validate the AI-generated code?

You can now see how properly somebody solved the issue and never simply the output.

Do algorithm questions nonetheless have worth? Yes, they do however the issue is the weightage related to them.