The interview process involved one hiring manager round, followed by a technical assessment (coding and system design short answers), and finally an on-site interview.

Unfortunately, they did not proceed with the on-site interview because the role had already been filled.

We are working on a clone of Facebook that allows users to share content with their social network. Users view posts in an infinite stream.

Our product only has a few thousand users, but we expect it to be as popular as Facebook very soon.

We want to add a live numeric count to every post showing how many friends the post's author has.

Right now, our database has two tables: 'user' and 'user_relationship'. Each friendship relation appears in the database exactly once.

The database definition includes:

• USER

  • id (primary key)
  • name
  • created_date

• USER_RELATIONSHIP

  • id (primary key, unique to each relationship)
  • user1_id (indexed)
  • user2_id (indexed)
  • start_date

What would you do with the database in order to provide this feature for use at a large scale?