There are two common coding methods used in technical interviews and assessments: algorithmic coding (e.g., Leetcode) and file system-based project coding. The latter is generally considered better because it is better suited for real-world work. In the current age of assisted programming tools (e.g., ChatGP and Github Copilot), there is less need for handwriting algorithms such as inverted binary trees. However, algorithmic coding is advantageous for its time-efficiency, making it ideal for time-critical interviews.
Based on this, our team has tried a simplified TDD coding approach in technical interviews, and it has proven successful and it strikes a balance between time-saving and job-oriented . We plan to release this method to the public, as we believe that TDD is not only a testing process but also a design process. Even if the candidate is unable to complete the coding challenge within the given time, the process of designing unit tests can demonstrate their coding skills. Additionally, the interviewer can quickly adjust the requirements by modifying the test cases.
Although the demo is currently simple, our goal is clear: to explore coding methods that balance time-saving and work-oriented principles in code evaluation, particularly in a time when programming tools are rapidly evolving.
What do you think about it?
