Merge is a new hiring platform that replaces Leetcode-style interviews with AI-native code review assessments. Candidates review a pull request against a small codebase, then an AI agent responds to their comments in real time, simulating a real engineer. The platform evaluates bug coverage, communication quality, PR revision count, and token use efficiency — metrics the founders argue are increasingly relevant as AI-generated code becomes the norm and engineering judgment matters more than code output.
Questions this post answers
How does the Merge code review hiring assessment work?
Candidates are shown a small codebase and a pull request to review. An AI agent responds to each comment in real time, simulating a real engineer, for up to 5 revision rounds. The platform then scores candidates on bug and vulnerability coverage, communication efficiency and constructiveness, and how many revisions and tokens the review consumed. Teams rethinking their engineering hiring process track tools like Merge on daily.dev.