New hiring tests challenge engineering grads to 'refactor' AI code

Following the recent hiring slowdown in the IT sector due to global economic challenges, fresh engineering graduates are now facing new types of coding assessments to secure jobs. According to a report in the Economic Times, entry-level and junior engineers are being tested in innovative ways: rather than creating code from scratch like in traditional hiring tests, they’re now asked to optimise and refine AI-generated code.

Fresh graduates seeking IT roles are now being asked to “refactor” AI-generated code as part of the recruitment process, rather than writing code from scratch. 

The Economic Times reports that “from being asked to write simple codes, candidates are being asked to optimise and restructure existing AI written codes, to improve quality.”

This shift reflects a growing focus in the industry on broader problem-solving skills over traditional coding. 

According to Chirag Mehta, Vice President and Principal Analyst at Constellation Research, hiring practices are now emphasising candidates’ ability to adapt, optimise, and enhance existing code for better performance.

“The focus in hiring has shifted from just programming skills to broader problem-solving capabilities,” Mehta told ET, adding, “Hiring tests for entry-level engineers are evolving to gauge candidates' ability to learn new methods, optimise existing code, and ‘refactor’ it to improve quality. It’s about finding engineers who are agile thinkers, not just scriptwriters.

Refactoring is the process of restructuring existing code without changing its external behavior or functionality. The goal is to improve the code’s readability, maintainability, and efficiency, making it easier to understand and work with. 

During refactoring, developers modify the internal structure of the code, optimise it, and remove redundancies, but the output or behavior of the program remains the same.

Yugal Joshi, a partner at Everest Group, added that as AI-driven tools become commonplace, developers will increasingly need skills beyond coding.

Joshi noted that the ability to assess AI-generated code and documentation—such as code comments and requirements—is now just as important as technical development skills.

For more information on exams, results and careers, please visit