Design Flight Booking and Voting Systems
Company: Hot Agent Startup
Role: Software Engineer
Category: System Design
Difficulty: medium
Interview Round: Onsite
Answer both system design prompts.
## Part A: Flight Ticket Booking System
Design the backend for a flight ticket booking product. The system should support searching flights, selecting an itinerary, booking tickets for passengers, and preventing overselling.
Cover the following:
- Data model for entities such as `Flight`, `Passenger`, and `FlightTicket` or `Booking`.
- Indexes to build and why.
- How B-tree indexes work internally and how that affects query design.
- End-to-end user experience flow and the backend queries for booking a flight.
- How to avoid double booking when many users try to buy the same flight or seat concurrently.
## Part B: Real-Time High-QPS Voting System
Design a real-time voting system for a fast-food promotion during a major live event. Users can quickly tap to vote for a taco flavor and should see a live results board.
Requirements:
- Traffic may reach a few million votes per second for a short burst.
- The vote action should have low latency.
- The results board should update in near real time.
- The design should be reliable under sudden traffic spikes.
Cover the following:
- APIs, services, storage, and data flow.
- How to use Kafka or another durable streaming system.
- How to model real-time counters and time-series metrics.
- Relevant Kafka internals and time-series database internals.
- Deployment strategy on Kubernetes or EKS, including autoscaling.
Overview: This question evaluates system design competencies including large-scale backend architecture, data modeling and indexing, database internals, concurrency control to prevent overselling, and high-throughput real-time streaming and metrics for live updates.