HackerRank is replacing its traditional coding tests with an artificial intelligence agent that interviews software engineering candidates directly, scoring their problem-solving process rather than just their final code. The platform, called Chakra, exits beta and becomes generally available to enterprise customers on Monday.
The launch marks a strategic reversal for the Y Combinator-backed company, which built a 30-million-developer network since its 2012 debut by selling standardized coding challenges to more than 3,000 employers including Amazon and Nvidia. CEO Vivek Ravisankar said generative AI has rendered output-based testing insufficient, because automated tools can now produce working code on demand. Chakra shifts the evaluation focus to how developers frame problems, direct AI assistants, and validate machine-generated solutions — a competency the company terms “AI fluency.”
During a Chakra session, candidates work inside a canvas featuring a real-world code repository alongside an embedded AI assistant. The interviewing agent monitors their workflow in real time, generating contextual follow-up questions about architectural choices or asking how a solution would adapt if new constraints were introduced. This format is designed to collapse three standard hiring phases — a recruiter screen, a take-home technical assessment, and a live engineering interview — into a single session.
Embedding an AI assistant directly into the testing environment produced an unexpected security outcome. HackerRank reported that suspicious-activity flags dropped between 70% and 80% during Chakra sessions compared to its legacy assessments, though variance persisted based on candidate seniority and geography. By officially permitting AI use, the system removes the incentive for candidates to covertly consult unauthorized external tools while being monitored.
More than 500,000 interviews were completed during the six-month beta period. Enterprise data firm Snowflake, AI company Snorkel, and IT services provider Capgemini participated in early testing alongside HackerRank’s own internal recruiting operations.
Chakra scores candidates against employer-defined rubrics but does not make final hiring decisions. Ravisankar positioned the tool as a mechanism to standardize technical evaluation, arguing that algorithmic consistency reduces the subjective bias human interviewers introduce when influenced by a candidate’s educational pedigree or background. However, applying identical criteria uniformly does not eliminate bias if the underlying training data or evaluation models contain historical disparities.
Automated hiring systems face mounting legal requirements across the United States. New York City’s Local Law 144 mandates that employers using automated employment decision tools conduct annual independent bias audits and notify candidates before deployment. Several other states, including Colorado, California, and Illinois, have advanced or enacted legislation governing AI in employment screening, requiring vendors like HackerRank to build compliance infrastructure directly into their platforms.
The broader technical assessment market is undergoing rapid consolidation as generative AI disrupts legacy screening methods. Competitors such as Codility, HackerEarth, and CodeSignal are similarly integrating AI-driven evaluation features, moving the industry away from timed algorithmic puzzles toward scenario-based testing that mirrors actual development environments. HackerRank intends to maintain its legacy testing products for clients that still require them, but Ravisankar identified Chakra as the company’s primary growth vector going forward.





