Design favorites and social game recommendations

Problem

Design a backend system for a gaming app that supports:

1. Favorites ("Fav") : users can favorite/unfavorite games and view their favorites list.
2. Social recommendations : show recommended games such as "Your friends played this game" / "Popular among your friends" .

Assume the product has tens of millions of users and a large game catalog.

Requirements

Functional

• Favorite/unfavorite a game.
• List a user’s favorited games (paginated, sorted by most recent).
• Generate a social recommendation feed for a user based on friends’ activity:
  • Friends who played a game recently
  • Friends who favorited/liked a game
  • Optional: trending in your friend network

Non-functional

• Low latency for reads (favorites list and recommendations).
• Correctness guarantees for favorites (no duplicates; idempotent operations).
• Handle privacy settings (e.g., user hides activity; blocked users).
• Scalable ingestion of play/favorite events.
• Reasonable freshness (e.g., within minutes) for recommendations.

Output

Describe APIs, data/storage design, event pipeline, recommendation logic/ranking, caching, and how you would scale and measure quality.