1. Object and blob storage

An instructor on an online course platform finishes recording lesson 812 and uploads a 2.4GB lecture video plus a slide deck as a PDF file. Thousands of students will fetch both. Storing large files in the platform's database would make backups and replication expensive.

Large files belong in an object store: a service such as Amazon S3 that keeps each file as an opaque blob of bytes under a name you choose, copies it across machines so it survives disk failures, and hands it back whole over HTTP. The database keeps a small row that says what the file is and where it lives.

Almost every prompt with files in it reaches this point: video platforms, photo apps, chat attachments, backups, logs. The segment is usually short, two to five minutes. At mid-level, putting the bytes in the store and the key in the row, then uploading straight from the client, is a strong answer. At senior level, interviewers push further: how a 4GB upload survives a dropped connection, why students still see old slides after an update, and how you list an instructor's files.

By the end you can split any file-heavy feature into rows and objects, describe how bytes get in and out, and name the limits the store imposes.