HackerRank's Chakra AI agent now runs job interviews and scores candidates

Chakra moves from beta to general availability

HackerRank, the platform companies use to assess and hire developers, has made its AI interviewing agent Chakra generally available to customers on Monday, according to TechCrunch. The agent conducts interviews, watches candidates as they work, and evaluates not only the answers they produce but the reasoning behind them. Chakra spent roughly six months in beta, during which HackerRank says it ran more than 500,000 interviews. Snowflake, Snorkel and Capgemini were among the companies that tested it, alongside HackerRank's own internal use.

How a Chakra interview works

Rather than a conventional coding test, a Chakra session is designed to resemble the job itself, TechCrunch reports. The candidate receives a task built around a real-world code repository and works through it in a canvas that includes an AI assistant. Chakra draws on the context of the candidate's work to ask follow-up questions mid-task, such as why one approach was chosen over another, or how the solution would change if a new constraint were introduced.

HackerRank co-founder and CEO Vivek Ravisankar frames the shift around what AI does to the value of finished output. The old evaluation model judged the artifact, he told TechCrunch, but now that "anybody can produce an artifact," employers need to understand the thinking and judgment that produced it. Beyond correctness, Chakra is meant to capture harder-to-measure signals such as critical thinking and what HackerRank calls "AI fluency": how well a candidate frames a problem for an AI system, judges its output, and steers it toward a solution.

Fewer flags and fewer rounds

HackerRank reports two notable effects. Suspicious-activity flags were 70 to 80 percent lower in Chakra interviews than in comparable traditional HackerRank assessments, though the rate varied with factors such as geography and seniority. Ravisankar's explanation is that when candidates are openly given an AI assistant, the incentive to secretly use outside tools that feed them answers largely disappears.

Chakra is also collapsing the hiring funnel. Ravisankar said the typical three-stage process — a recruiter screen, a take-home assessment and a follow-up interview with an engineer — can now be merged into a single Chakra interview.

A bet against HackerRank's old business

The product effectively undercuts the assessment business HackerRank has built since launching at TechCrunch Disrupt in 2012. Its traditional offering largely tested whether developers solved coding problems correctly, and Ravisankar argues AI has made that a weaker signal of engineering ability. He compared the transition internally to Apple moving from the iPod to the iPhone, with the old product retaining value but the new one pointing where the market is headed. "Chakra is going to be the headline," he said. The Y Combinator-backed company counts more than 3,000 business customers, including Amazon, Nvidia, Clay and Replit, and a community of over 30 million developers.

Bias and regulatory questions

Giving an algorithm a deeper role in judging people raises its own issues. Ravisankar said Chakra is designed to score candidates rather than make the final hiring decision, which stays with humans, and argued that an AI system can be instructed to follow the same rubric for every candidate instead of being swayed by background or education.

TechCrunch points out that applying criteria consistently does not make an automated tool bias-free: such systems can inherit or amplify bias from the data, models and criteria used to build them, which is one reason regulators are scrutinising them. New York City, for example, requires employers using certain automated employment decision tools to submit them to an independent bias audit and notify candidates beforehand. Ravisankar acknowledged that hiring is a regulated area and said compliance has been part of what HackerRank had to build for.

Why it matters

Chakra is a concrete example of AI moving from assisting candidates to evaluating them, and at meaningful scale given the beta volume. The change in what gets measured — process and judgment, including how well someone directs AI, rather than final output — could reshape technical hiring well beyond HackerRank's customer base. At the same time, the question of how much of a hiring decision should be delegated to an algorithm, and whether consistent scoring truly reduces bias, has stopped being hypothetical and is now a practical and regulatory concern for thousands of companies.