HackerRank launches AI agent to conduct and evaluate developer interviews

HackerRank, a platform used by companies to assess and hire developers, launched Chakra, an AI agent designed to conduct interviews, observe candidates as they work, and evaluate their problem-solving process rather than just their final answers. The product has completed over 500,000 interviews during its six-month beta period, with early customers including Snowflake, Snorkel, and Capgemini, as well as internal testing at HackerRank.

Chakra’s interviews are structured to resemble real work by presenting candidates with tasks involving actual code repositories. Candidates work in an interface that includes an AI assistant, and the AI interviewer asks follow-up questions about their approach, such as why they chose a particular method or how their solution would adapt to new constraints. After the interview, Chakra generates a detailed report for hiring managers.

The company claims that allowing candidates to use AI during interviews reduced suspicious-activity flags by 70% to 80% compared to traditional HackerRank assessments, attributing this to a lower incentive for candidates to secretly use external tools that feed them answers.

HackerRank’s traditional assessment model focused on whether developers could solve coding problems correctly, but the company now argues that AI has made this approach less effective for measuring engineering ability. Chakra is positioned as the future of HackerRank’s product, combining what previously required three separate hiring rounds—recruiter screen, take-home assessment, and engineer interview—into a single session.

HackerRank co-founder and CEO Vivek Ravisankar stated that Chakra is designed to score candidates rather than make final hiring decisions, which remain with humans. He argued that AI can handle structured evaluation more consistently than humans, though he acknowledged that AI systems can still inherit or amplify biases if not properly tuned.