Open Book Exams and Coding Interviews

When I was in university, sometimes professors gave open book exams, so I could search for complex logic or functions to solve the questions. Still, for most exams, we needed to memorize everything and try to anticipate the questions we would be given.

After graduation, I joined a company as a software engineer. Back then, I had no clue how to code well; I simply tried my best, read big red books, and looked for tutorials. Now that I am a hiring manager, I often have to find a way to assess candidates. One of the current trends is the so-called LC (LeetCode), which refers to brain teasers regarding algorithms, data structures, arrays, and sorting. Regardless of how skilled you are, job seekers often need to dedicate personal time to studying this.

Typically, companies provide random questions and ask candidates to solve them within a limited time frame. Some well-known (big) companies use online platforms for coding exams. While this approach may seem logical, it reminds me of university exams. To excel in those tests, I had to focus on specific areas and hope that the questions I anticipated would be asked. If they weren’t, I would receive lower grades. However, that doesn’t mean someone is unintelligent.

The real problem lies in the fact that while we ask these questions during the hiring process, in most companies or software engineering roles, we seldom use them in our day-to-day work. This is especially true in startups, where there is pressure from CXOs to release product features quickly, leaving little time for certain tasks. Engineers often need to search for solutions through online resources like Google, Stack Overflow, and even AI like ChatGPT.

The reality in most engineering roles resembles an “Open Book Exam.” So why do we consistently seek individuals who excel in areas we may rarely use in different skill sets? It is clear that we need to find a much better way to assess engineers instead of relying on this peculiar approach.

#startups #hiring #coding