Case Study: Design a News Feed Ranking System

Mental Model

Connecting isolated components into a resilient, scalable, and observable distributed web.

How does TikTok or Facebook decide what you see next? It's a combination of High-Scale Distributed Systems and Real-Time Machine Learning Scoring.

1. Requirement Clarification

graph LR
    Producer[Producer Service] -->|Publish Event| Kafka[Kafka / Event Bus]
    Kafka -->|Consume| Consumer1[Consumer Group A]
    Kafka -->|Consume| Consumer2[Consumer Group B]
    Consumer1 --> DB1[(Primary DB)]
    Consumer2 --> Cache[(Redis)]

Functional

• Suggest relevant content to users.
• Update recommendations in real-time as users interact.
• Handle massive throughput (millions of users).

Non-Functional

• Latency: Sub-100ms response time.
• Personalization: Unique feed for every user.
• Scalability: Handle billions of items and users.

2. Recommendation Architecture

1. Candidate Generation (Retrieval): Fast filtering of billions of items down to ~1,000 potential candidates (using Vector Search/Collaborative Filtering).
2. Scoring (Ranking): Using a deep ML model to predict the probability of a user clicking/liking each candidate.
3. Re-ranking: Applying business logic (e.g., deduplication, diversity, ad placement).

3. The Cold Start Problem

How do you recommend items to a brand-new user?

• Use Popularity-based or Demographic-based content.
• Ask for user interests on signup.

4. Scaling the Scoring Layer

Scoring is computationally expensive.

• Fix: Use Model Pruning and Batching. Cache common recommendation results if possible.

Final Takeaway

Ranking systems are a balance of Relevance vs. Latency. The best model is useless if it takes 5 seconds to load the feed.

Engineering Standard: The "Staff" Perspective

In high-throughput distributed systems, the code we write is often the easiest part. The difficulty lies in how that code interacts with other components in the stack.

1. Data Integrity and The "P" in CAP

Whenever you are dealing with state (Databases, Caches, or In-memory stores), you must account for Network Partitions. In a standard Java microservice, we often choose Availability (AP) by using Eventual Consistency patterns. However, for financial ledgers, we must enforce Strong Consistency (CP), which usually involves distributed locks (Redis Redlock or Zookeeper) or a strictly linearizable sequence.

2. The Observability Pillar

Writing logic without observability is like flying a plane without a dashboard. Every production service must implement:

• Tracing (OpenTelemetry): Track a single request across 50 microservices.
• Metrics (Prometheus): Monitor Heap usage, Thread saturation, and P99 latencies.
• Structured Logging (ELK/Splunk): Never log raw strings; use JSON so you can query logs like a database.

3. Production Incident Prevention

To survive a 3:00 AM incident, we use:

• Circuit Breakers: Stop the bleeding if a downstream service is down.
• Bulkheads: Isolate thread pools so one failing endpoint doesn't crash the entire app.
• Retries with Exponential Backoff: Avoid the "Thundering Herd" problem when a service comes back online.

Critical Interview Nuance

When an interviewer asks you about this topic, don't just explain the code. Explain the Trade-offs. A Staff Engineer is someone who knows that every architectural decision is a choice between two "bad" outcomes. You are picking the one that aligns with the business goal.

Performance Checklist for High-Load Systems:

1. Minimize Object Creation: Use primitive arrays and reusable buffers.
2. Batching: Group 1,000 small writes into 1 large batch to save I/O cycles.
3. Async Processing: If the user doesn't need the result immediately, move it to a Message Queue (Kafka/SQS).

Technical Trade-offs: Messaging Systems

Key Takeaways

• Suggest relevant content to users.
• Update recommendations in real-time as users interact.
• Handle massive throughput (millions of users).

Read Next

• Case Study: Design a Real-Time Analytics System
• Deep Dive: Vector Embeddings

Verbal Interview Script

Interviewer: "How would you ensure high availability and fault tolerance for this specific architecture?"

Candidate: "To achieve 'Five Nines' (99.999%) availability, we must eliminate all Single Points of Failure (SPOF). I would deploy the API Gateway and stateless microservices across multiple Availability Zones (AZs) behind an active-active load balancer. For the data layer, I would use asynchronous replication to a read-replica in a different region for disaster recovery. Furthermore, it's not enough to just deploy redundantly; we must protect the system from cascading failures. I would implement strict timeouts, retry mechanisms with exponential backoff and jitter, and Circuit Breakers (using a library like Resilience4j) on all synchronous network calls between microservices."