Netflix

Design a COVID-19 contact tracing system with the following features:

Requirements:

1. Mobile Device Logging

• Maintain 2-week rolling log of encounters
• Record timestamps and device identifiers
• Efficient storage management

2. Data Retrieval & Processing

• Use Cassandra for log storage
• Implement BFS for contact network traversal
• Handle large-scale data efficiently

3. System Architecture

• Mobile device encounter logging
• Central database management
• Contact network analysis

Example Scenario: When User A reports COVID:

1. Retrieve User A's contact logs
2. Perform BFS to identify direct contacts
3. Continue traversal for indirect exposure
4. Notify affected users

Find the shortest path between two products in a warehouse.

Implementation Requirements:

1. Data Model

• Use graph representation
• Nodes represent locations
• Edges represent pathways
• Support for hallways and rooms

2. Algorithm

• Implement BFS for shortest path
• Return path as array
• Handle invalid paths

Example:

Input:
- Warehouse graph
- Start product: A
- End product: B

Output: Array representing shortest path from A to B

Design a system for video downloads and creator subscriptions.

Key Features:

1. Video Downloads

• Purchase individual videos
• Secure download access
• Content delivery system

2. Creator Subscriptions

• Subscription management
• Access control
• Content updates

3. Payment Processing

• Secure transactions
• Refund handling
• Purchase history

4. System Requirements

• User authentication
• Scalable architecture
• Transaction management
• Content delivery optimization

Interview experience from a Google L4 Infrastructure Engineer:

Phone Screen:

• Interviewed by Holden Karau (Spark expert)
• Mostly career discussion and Netflix culture
• Brief 10-minute technical question
• Very conversational approach

Virtual Onsite (Day 1):

1. Concurrency Round

• Focus on mutex/condition variables
• Basic operations and trade-offs
• Emphasis on clear communication

2. Problem Solving Round

• LRU Cache variant
• Leetcode-style implementation
• Moderate difficulty level

3. System Design Round

Topic: Scalable Counter System
Requirements:
- Start with single server, bare metal
- Scale up to cloud architecture
- Handle view counts/metrics
- Deep dive into implementation details

Key Focus Areas:
- Open-source tech stack choices
- Partition/replication algorithms
- Component-level deep dives
- Architectural assumptions

Process ended after Day 1.

Key Takeaways:

• Very professional interviewers
• High technical bar
• Deep system design knowledge required
• Important to understand fundamentals beyond existing platforms
• Recommended prep: "Designing Data-Intensive Applications"