Why Software Engineering Interviews are Broken and How to Actually Make them Better

Our process for assessing software engineers is fundamentally but not irrevocably broken. Currently the system fails in its most basic goal: identifying engineers who will succeed on the job. It disenfranchises both candidates and employers, screens out capable engineers, wastes time and resources, and undermines the hiring process itself.

We will first discuss the current landscape of software engineering interviews - including the most ubiquitous categories of interview problems used today. We will discuss how competitive programming - once a niche, recreational activity - has become a prerequisite for job placement for essentially all engineering roles.

From there, we will analyze the gap between what we are trying to assess with these problems and the actual skills that correlate with job performance.

We will discuss the core purpose of the interview process - minimizing risk in hiring by predicting future job performance - in order to understand why both engineers and the companies that employ them are poorly served by the current system. Then we will identify the most important indicators of high job performance and how we can best identify those skills in candidates.

Finally, we’ll present a practical framework for designing interview questions that surface these skills more effectively. Attendees will leave with actionable strategies for structuring high-signal, real-world interview processes that better predict impact and performance.