Hire better, faster.
Practical guides on developer hiring, take-home review, and AI-coding assessment.
- ·1 min read·
We tried using AI to detect AI-generated code. It didn't work.
Why we scrapped LLM-based detection for code submissions and went back to basics with regex patterns and git signals.
Read guide - ·8 min read·
How to detect AI-written code in technical interviews and take-home assignments
AI-generated code is increasingly hard to spot by eye. Here's what the signals actually are, which detection methods work at scale, and how to handle it when you find it.
Read guide - ·6 min read·
How to write a take-home coding assignment that actually predicts job performance
Most take-home assignments test the wrong things, take too long, and drive away good candidates. Here's how to write one that's fair, fast to review, and predicts day-one performance.
Read guide - ·7 min read·
ChatGPT vs CodeVerdict for reviewing take-home assignments: an honest comparison
ChatGPT can absolutely review code. Whether it's the right tool for screening 50 take-home submissions a week is a different question. Here's what each does well, what each gets wrong, and how to decide.
Read guide - ·3 min read·
How to review developer take-home assignments without losing your evening
Reviewing a single take-home assignment takes 15–30 minutes on average — and most of that time is spent on tasks AI can do in 60 seconds. Here's a faster, fairer rubric.
Read guide