Nested Comments & Collaboration APIs

Postman's coding round hands you a build. You open your own editor, stand up a local server against a database of your choice, and ship working REST endpoints in about an hour. The interviewer then hits those endpoints from a Postman collection and watches what comes back: the right status codes, sensible headers, clean validation on the requests that should fail. The nested comments service is the problem candidates draw most often for this round, so build it until the data model and the endpoints are automatic.

The job the round simulates is ordinary product work. A comment can reply to another comment, which can reply to another, so you are storing a tree in a flat table and reading it back as a thread. None of that needs a clever algorithm. What it needs is a data model that holds up, endpoints that behave, and validation that rejects the bad request with the right code instead of a 500. That is exactly what the round scores, and it is where most of this lesson lives.

One thing to set expectations. Data-structures-and-algorithms questions do show up at Postman, but only as a light phone-screen filter, easy-to-medium and quick. One candidate who cleared the loop put it plainly: no DSA was asked, and it was a breath of fresh air. The round that decides the outcome is this one, the machine-coding build. If your invite mentions setting up a local IDE, a server, and a database beforehand, this is the round you are walking into.