##### Question Design a news-feed feature for a food discovery app (DoorDash-style). The feed lists items showing **title, author, short summary, and engagement counts (comments, likes/saves)**, and tapping an item opens a **detail view** with the full description, comments, and related items. Cover the full system end to end: 1. **Data model & storage.** Define the entities (posts, comments, reactions/likes/saves, users, follows if needed, materialized counters, and a per-user or per-segment feed). Specify the storage choice for each (relational core vs. wide-row NoSQL vs. Redis) and the denormalized/materialized views you keep. 2. **APIs & pagination.** Read/write endpoints: list feed, fetch post detail, create/update post, create/update comment, toggle reaction, and search/filter. Use cursor-based (keyset) pagination and infinite scroll. Show the request/response shape and how cursors are encoded. 3. **Ranking.** Rank by recency + engagement. Give a concrete scoring formula with time decay and explain how scores are updated as engagement arrives. 4. **End-to-end request path.** Trace a request after it enters the system: edge/CDN, API gateway (auth, rate limiting), the stateless app tier, caches, OLTP/NoSQL stores, background workers, and the message queue / event bus. Do this for the **feed read**, the **post-detail read**, the **comment/reaction write**, and the **new-post create path**. 5. **New-post propagation.** Describe how a newly created post is processed and propagated into user feeds — including a fan-out strategy (fan-out-on-write vs. fan-out-on-read, and how you handle very-high-follower / celebrity authors). 6. **Caching, indexing & denormalization** for fast reads (hot-feed caches, counter caches, search index, author-snippet denormalization, hot-key protection). 7. **Search & filter** (e.g. by cuisine, city, tags, author) backed by a search engine. 8. **Scalability & resilience.** Partitioning/sharding, rate limiting, backpressure, circuit breakers, and graceful degradation. 9. **Consistency & idempotency.** Which paths are strongly vs. eventually consistent; read-your-writes for authors; idempotency keys for writes; idempotent counter aggregation. 10. **Error handling, retries, observability, and performance optimizations** (latency targets, batching, connection pooling, payload trimming, reconciliation jobs). State your assumptions and rough traffic estimates, and call out the key trade-offs.

A complete system-design walkthrough for a DoorDash-style food news feed: data model (posts, comments, reactions, follows, materialized feeds), cursor-paginated REST APIs, a recency-plus-engagement ranking formula, and the full end-to-end request path through CDN, gateway, app tier, caches, OLTP/NoSQL stores, and a Kafka event bus. Covers hybrid fan-out propagation for new posts, caching and denormalization, search/filter, partitioning, rate limiting and backpressure, consistency and idempotency, observability, and performance optimizations.

How do I approach System Design interview questions?

System Design questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master system design interviews.

What difficulty level is this interview question?

This is a hard difficulty System Design question, commonly asked during Onsite rounds at DoorDash.

What role is this question designed for?

This question is commonly asked for Software Engineer candidates at DoorDash during technical interviews.

Design a scalable food news feed | DoorDash Interview Question

Question

Design a news-feed feature for a food discovery app (DoorDash-style). The feed lists items showing title, author, short summary, and engagement counts (comments, likes/saves), and tapping an item opens a detail view with the full description, comments, and related items. Cover the full system end to end:

Data model & storage. Define the entities (posts, comments, reactions/likes/saves, users, follows if needed, materialized counters, and a per-user or per-segment feed). Specify the storage choice for each (relational core vs. wide-row NoSQL vs. Redis) and the denormalized/materialized views you keep.
APIs & pagination. Read/write endpoints: list feed, fetch post detail, create/update post, create/update comment, toggle reaction, and search/filter. Use cursor-based (keyset) pagination and infinite scroll. Show the request/response shape and how cursors are encoded.
Ranking. Rank by recency + engagement. Give a concrete scoring formula with time decay and explain how scores are updated as engagement arrives.
End-to-end request path. Trace a request after it enters the system: edge/CDN, API gateway (auth, rate limiting), the stateless app tier, caches, OLTP/NoSQL stores, background workers, and the message queue / event bus. Do this for the feed read , the post-detail read , the comment/reaction write , and the new-post create path .
New-post propagation. Describe how a newly created post is processed and propagated into user feeds — including a fan-out strategy (fan-out-on-write vs. fan-out-on-read, and how you handle very-high-follower / celebrity authors).
Caching, indexing & denormalization for fast reads (hot-feed caches, counter caches, search index, author-snippet denormalization, hot-key protection).
Search & filter (e.g. by cuisine, city, tags, author) backed by a search engine.
Scalability & resilience. Partitioning/sharding, rate limiting, backpressure, circuit breakers, and graceful degradation.
Consistency & idempotency. Which paths are strongly vs. eventually consistent; read-your-writes for authors; idempotency keys for writes; idempotent counter aggregation.
Error handling, retries, observability, and performance optimizations (latency targets, batching, connection pooling, payload trimming, reconciliation jobs).

State your assumptions and rough traffic estimates, and call out the key trade-offs.

Question

Data model & storage. Define the entities (posts, comments, reactions/likes/saves, users, follows if needed, materialized counters, and a per-user or per-segment feed). Specify the storage choice for each (relational core vs. wide-row NoSQL vs. Redis) and the denormalized/materialized views you keep.
APIs & pagination. Read/write endpoints: list feed, fetch post detail, create/update post, create/update comment, toggle reaction, and search/filter. Use cursor-based (keyset) pagination and infinite scroll. Show the request/response shape and how cursors are encoded.
Ranking. Rank by recency + engagement. Give a concrete scoring formula with time decay and explain how scores are updated as engagement arrives.
End-to-end request path. Trace a request after it enters the system: edge/CDN, API gateway (auth, rate limiting), the stateless app tier, caches, OLTP/NoSQL stores, background workers, and the message queue / event bus. Do this for the feed read , the post-detail read , the comment/reaction write , and the new-post create path .
New-post propagation. Describe how a newly created post is processed and propagated into user feeds — including a fan-out strategy (fan-out-on-write vs. fan-out-on-read, and how you handle very-high-follower / celebrity authors).
Caching, indexing & denormalization for fast reads (hot-feed caches, counter caches, search index, author-snippet denormalization, hot-key protection).
Search & filter (e.g. by cuisine, city, tags, author) backed by a search engine.
Scalability & resilience. Partitioning/sharding, rate limiting, backpressure, circuit breakers, and graceful degradation.
Consistency & idempotency. Which paths are strongly vs. eventually consistent; read-your-writes for authors; idempotency keys for writes; idempotent counter aggregation.
Error handling, retries, observability, and performance optimizations (latency targets, batching, connection pooling, payload trimming, reconciliation jobs).

State your assumptions and rough traffic estimates, and call out the key trade-offs.

Design a scalable food news feed

Quick Overview

Question

Solution

Submit Your Answer to Earn 20XP

Design a scalable food news feed

Quick Overview

Question

Solution

Submit Your Answer to Earn 20XP