System Design: Cheapest Flight Search Across Multiple Providers

Context

Design a backend system that searches multiple flight providers (airlines, OTAs, GDS/aggregators) to return the cheapest flight tickets for users. The system must handle a high volume of queries, volatile prices, and frequent updates while remaining scalable and cost-efficient.

Requirements

• Inputs: origin, destination, departure date(s), return date(s), cabin class, number of stops.
• Date flexibility: exact dates or flexible windows (e.g., ±3 days) and calendar view.
• Core outputs: list of itineraries/offers sorted by price with filters, plus a fast “cheapest per day” calendar.
• System concerns to describe:
  1. High-level architecture: ingestion from providers, normalization, deduplication, search index.
  2. APIs: search, quote/validate (re-price), and ancillary endpoints.
  3. Caching strategy.
  4. Data freshness and update propagation.
  5. Rate limiting, retries, and resilience patterns.
  6. Consistency vs availability trade-offs.
  7. Scalability and cost-control strategies.
  8. Schemas and indexes that enable fast queries and rapid price updates.