System Design: Worked Examples

Five detailed system design examples with complete API design, data models, architecture, failure handling, and trade-offs.

Examples

1. URL Shortener → 01-URL-Shortener.md

Problem: Convert long URLs to short codes; redirect with analytics
Scale: 1M shorten requests/day, 100M redirects/day
Key Topics: Base62 encoding, caching, distributed ID generation, analytics
Duration in interview: 45 minutes

2. News Feed (Social Media) → 02-News-Feed.md

Problem: Personalized feed ranking posts in real-time
Scale: 1B users, 500M DAU, 50M posts/day
Key Topics: Push vs. Pull models, ML ranking, cache strategies, sharding
Duration in interview: 60 minutes

3. Chat System (Real-Time Messaging) → 03-Chat-System.md

Problem: 1-to-1 and group messaging with delivery guarantees
Scale: 100M DAU, 1B messages/day, 50M concurrent connections
Key Topics: WebSocket, message queuing, offline sync, ordering guarantees
Duration in interview: 60 minutes

4. Rate Limiter → 04-Rate-Limiter.md

Problem: Protect APIs from abuse; limit requests per user/IP
Scale: 100K requests/sec to rate limiter
Key Topics: Token bucket, sliding window, Redis, multi-level limiting
Duration in interview: 30–45 minutes

5. File Storage & CDN → 05-File-Storage-CDN.md

Problem: Durable file storage with global distribution and low latency
Scale: 1B uploads/day, 100B downloads/day
Key Topics: Multi-part upload, CDN, replication, deduplication, tiering
Duration in interview: 60 minutes

How to Use These Examples

During Interview Prep

Read one example per day (takes 20–30 min)
Then design it yourself from scratch (takes 45–60 min)
Compare your design against the worked solution
Iterate: What did you miss? What would you do differently?

During Interview

Use these as templates, not scripts
Adapt to the specific prompt you receive
Follow the structure: Clarify → High-level → Deep dive → Tradeoffs → Failure handling
Discuss with interviewer as you go; don't just present

Common Variations (For Practice)

URL Shortener: "Add real-time trending URLs" / "Support custom expiration"
News Feed: "Add video support" / "Handle 100M concurrent users"
Chat: "Add end-to-end encryption" / "Support video calls"
Rate Limiter: "Add weighted limits" / "Support distributed rate limiting"
File Storage: "Add versioning" / "Support live streaming"

Interview Strategy

Your Goal

Demonstrate you can design scalable systems
Show communication + collaboration with interviewer
Prove you understand tradeoffs, not just technologies

Success Metrics (Use 05-Mock-Interviews/01-Self-Evaluation-Rubric.md)

Correctness (5/5): Design handles the scale; no critical flaws
Communication (5/5): Explain clearly; ask interviewer for feedback
Tradeoffs (5/5): Discuss 2–3 key tradeoffs; justify decisions
Failure handling (5/5): Address 2–3 failure scenarios

Common Interview Questions (Based on These Examples)

Depth Questions

"How does your database scale to 1B records?" → Sharding strategy
"What if cache goes down?" → Fallback to origin; latency degrades
"How do you prevent message loss?" → Replication, durability guarantees

Variant Questions

"What if we need 10x scale?" → Add more nodes, sharding, multi-region
"How do we save costs?" → Compression, tiering, deduplication
"How do we ensure high availability?" → Redundancy, failover, monitoring

Tradeoff Questions

"Consistency vs. Availability?" → Depends on use case; discuss both
"Latency vs. Cost?" → Cache for fast, archive for cheap
"Simplicity vs. Scale?" → Start simple, add complexity when needed

Study Path

Beginner (Weeks 1–2): Study all 5 examples
Intermediate (Weeks 3–4): Design each from scratch without looking at solution
Advanced (Weeks 5–6): Combine ideas (e.g., "Design Twitter" = feed + real-time + storage)
Expert (Weeks 7–8): Receive novel prompts; improvise using patterns learned

These examples cover ~70% of real FAANG system design interviews. ✅