AI is becoming exponentially better everyday. At Sardine it is an essential part of our day-to-day development activity. This means a leetcode-style coding interview is no longer the optimal way of interviewing candidates. Here's how we're updating our interview process to better evaluate AI-assisted real-world coding skills in 2025.

Limitation of leetcode-on-coderpad approach

The traditional algorithm-heavy technical interview faces two significant challenges:

1. AI has transformed solution discovery - Coding assistants can quickly generate solutions to standard algorithm problems.
2. Traditional interviews don't reflect modern workflows - Developers routinely use AI tools like Claude and GitHub Copilot, making isolated coding exercises less representative of day-to-day work.

While we still need to assess coding ability, we believe the evaluation should focus on how effectively candidates can solve problems using all available resources—including AI.

Our New Technical Interview Format

We've been trying an interview format that reflects real-world settings

• Candidates join with their preferred development environment - Bring your own IDE, tools, and AI coding assistants.
  • Use of search engines and AI assistants is encouraged, reflecting real-world development workflows.
• We start with a simple foundation - we provide basic server implementation in various languages.
• The interview features real-world problems - Tasks that mirror actual development scenarios rather than algorithm puzzles.
  • For example, candidates are given an API to accept a list of payment transactions and return the risk score. We then work together to implement the risk scoring logic - for instance, we check the total amount of transactions, or we call an external API (a mock prepared by us) to get additional data.
• Iterative problem-solving - As candidates progress, we adapt requirements.For example, we provide different test data, or we now ask candidates to implement different algorithm. This reflects real-world environment where specifications are constantly evolving.

Candidates share their entire screen during the process so we can work together while observing how they approach the problem.

Embracing the Future

We hope this approach creates a better assessment environment that:

• Acknowledges AI as a productivity multiplier rather than a threat.
• Evaluates how candidates operate in conditions similar to our actual workplace.
• Focuses on the human communication and problem-solving skills that AI can't replace.

Hiring great talent is essential to our success. We’re continuously iterating the interview process based on new technologies that become available to developers.

We are always looking for great people to join us. See our Careers page and apply today!

Also, we’d love to get your feedback. What kind of interview have you found useful (either as interviewer or interviewee)? We would love to hear from you! You can find me on linkedin.