Conducted 450+ technical interviews but couldn't find suitable candidates

Frontend
1

Our company recently advertised multiple positions on LinkedIn for entry-level frontend developers, backend engineers, and QA testers. The compensation package was competitive, going up to ₹20 lakhs annually. The response was overwhelming with over 12,000 applications flooding in.

We had to screen out approximately 10,000 applicants right away because their technical skills didn’t match our requirements or their CVs weren’t relevant to what we needed. We weren’t trying to be mean, we just wanted to save everyone’s time instead of putting unqualified people through lengthy interview processes.

During our technical assessments, we tested candidates on fundamental programming concepts and data structure algorithms including binary trees, priority queues, doubly linked lists, breadth-first search, depth-first search, and similar topics. We even permitted applicants to use ChatGPT for problem solving. But when we asked them to explain the time complexity or space complexity of their solutions, or simply describe what their code actually does, most couldn’t give proper answers.

Many applicants seem to be just copying and pasting AI-generated code without actually comprehending how it works. This trend is making it really tough to identify developers who genuinely know their stuff.

We’re beginning to wonder if there’s something wrong with our interview approach, or if junior developers really need to focus more on understanding code instead of just copying it from AI tools.

2

hmm curious - what if you flip the interview and start with a small project they’ve built instead of jumping into complex algos? Even basic stuff shows whether they actually understand their work or just pieced things together without knowing why.

3

yea totally agree! fresh grads are still figuring stuff out, expecting them to ace algo tests is tough. maybe try more hands-on stuff like pair programming, helps to see their thought process and how they approach problems in real-time.

4

Your screening looks solid, but the problem might run deeper than how you interview. Bootcamps and coding programs now push rapid deployment over actually understanding the basics. I hit the same wall when hiring at my last startup. What helped was starting candidates with dead simple problems - basic array stuff or string operations - then ramping up complexity. This separated people who got the fundamentals from those just memorizing patterns. Try having candidates explain their approach before they touch any code. Real developers will walk you through their logic, but anyone leaning on AI will fumble when they can’t articulate a strategy upfront.