Coupang - Senior Software Engineer Interview

Overview

Interviewed with Coupang for a Senior Software Engineer position. The process consisted of 3 comprehensive rounds covering DSA, High-Level Design, and Low-Level Design.

Interview Format: Virtual (Video Call) Total Rounds: 3 Duration: 2 days

---

Round 1: DSA - Maximum Window Subarray

Date: Tuesday, September 17, 2024 Duration: ~1 hour

Problem

Find the maximum sum of a subarray where:

• Subarray length ≤ k
• Handle positive and negative numbers
• Return maximum sum and indices

Approach: Sliding window technique

Complexity: O(n) time, O(1) space

Key Learning: Sliding window is essential for subarray problems

---

Round 2: High-Level Design - Job Scheduler

Date: Wednesday, September 18, 2024 Duration: ~1 hour

Problem

Design a distributed job scheduler for e-commerce platform.

Requirements:

• Schedule one-time and recurring jobs
• Support millions of jobs
• High availability and fault tolerance
• Priority-based execution
• Job dependencies
• Retry logic

Key Components:

• Job API Server (load balanced)
• Job Metadata DB (PostgreSQL)
• Job Queue (Kafka/RabbitMQ)
• Job Scheduler (with leader election)
• Job Executor (worker pool)
• Status Store (Redis)

Scale:

• 10M jobs scheduled
• 1000 jobs/second execution
• ~100GB storage with indexes

Key Decisions:

• Distributed scheduler with leader election
• Push model with backpressure
• At-least-once execution (idempotent jobs)

---

Round 3: Low-Level Design - Attendance Management

Date: Wednesday, September 18, 2024 Duration: ~1 hour

Problem

Design an attendance management system for warehouse workers.

Requirements:

• Multiple warehouses
• Different worker types (full-time, part-time, contract)
• Shift management (morning, evening, night)
• Check-in/check-out with location verification
• Attendance reports for payroll
• Overtime tracking

Core Classes:

• Worker (with FullTimeWorker, PartTimeWorker, ContractWorker)
• Warehouse
• Shift
• AttendanceRecord
• AttendanceSystem

Key Features:

• Location-based check-in validation
• Automatic overtime calculation
• Daily and period-based reporting
• Thread-safe operations

Design Patterns Used:

• Enum Pattern (worker types, shift types)
• Strategy Pattern (hours calculation by worker type)
• Repository Pattern (managing warehouses and records)

---

Key Takeaways

Preparation Time: 2-3 months

Focus Areas:

• Sliding window and array algorithms
• Distributed system design (job schedulers, queues)
• Object-oriented design principles
• E-commerce domain knowledge

Interview Style:

• Professional and structured
• Collaborative problem-solving
• Focus on practical solutions
• Balance of breadth and depth

Tips:

• Master sliding window pattern
• Practice system design for job/task schedulers
• Review OOP principles and design patterns
• Study e-commerce workflows
• Think about scale (millions of users/orders)
• Consider peak times (flash sales)