[withdavidli]

I think this is a smart way to scale up an applicant pool with minimal resources. My opinion is based on being in recruiting and doing analytics from startups to big tech.

It might be worse than former TripleByte interviews in figuring out signals, but it is much less people and cost intensive to scale. I think it's a smart trade off for a start up.

Most first interviews cover the same set of questions. For engineers, it's usually can this person actually code / leet code knowledge.

On-site interviews have a lot of variables, ranging from types of questions, interviewers, and hiring committee for an offer to be extended. I can see why being rejected at 2nd/on-site stage doesn't matter much.