HackerRank, the technical assessment platform, is pivoting its interview paradigm with the general release of Chakra, an AI agent capable of conducting full interviews. After a six-month beta period that included trials by companies such as Snowflake, Snorkel, and Capgemini, the tool officially launched on Monday. During testing, the system processed over 500,000 interviews, including sessions conducted internally at HackerRank.
Chakra moves beyond simply grading whether a candidate provides the correct answer. Instead, it evaluates the problem-solving process, measuring factors such as critical thinking, judgment, and what the company terms “AI fluency.” This fluency refers to a candidate’s ability to effectively prompt AI, assess its output, and guide it toward a solution.
“The previous modality of evaluation was evaluating the output,” said Vivek Ravisankar, co-founder and CEO of HackerRank. “Now, because of AI, anybody can produce an artifact.” He noted that the focus must shift to understanding the reasoning behind the work. In practice, candidates tackle tasks using real-world code repositories within an interface that includes an integrated AI assistant. As they work, Chakra observes their progress and asks follow-up questions about their approach or how constraints might alter their solution.
Ravisankar described the change as a consolidation of the traditional hiring funnel. Where candidates previously endured separate recruiter screens, take-home assessments, and engineering interviews, Chakra combines these into a single session. The platform also generated detailed reports for hiring managers after each interview.
Contrary to concerns that allowing AI access might encourage cheating, HackerRank reported that suspicious activity flags dropped by 70% to 80% compared to traditional assessments. Ravisankar attributed this to the fact that when AI is provided as a legitimate tool, the incentive to secretly seek outside aids diminishes.
The launch represents a significant evolution for the Y Combinator-backed startup, which was introduced at TechCrunch Disrupt in 2012. Originally built around coding challenges, HackerRank now serves more than 3,000 business customers, including Amazon and Nvidia, and hosts a community of over 30 million developers. Ravisankar likened the transition to Apple’s move from the iPod to the iPhone, acknowledging that while the traditional coding tests still hold value, the new model reflects where the market is heading.
Despite the technological shift, the company emphasized that Chakra is designed to score candidates rather than make final hiring decisions, which remain in human hands. Ravisankar argued that properly tuned AI can reduce bias by applying consistent rubrics, unlike human interviewers who may be influenced by background or education. However, experts note that automated tools can still inherit biases from their training data, prompting regulatory scrutiny. Laws in jurisdictions like New York City already require bias audits for certain automated employment decision tools, a compliance requirement HackerRank says it has integrated into its development process.
Leave a Reply