Case Study

How We Built a Multi-Provider Travel Booking Platform with FastAPI & Next.js

VViznetic EngineeringMay 20, 202612 min read

India's travel market is enormous and deeply fragmented. When we set out to build BookTravel, the goal was simple to state and hard to deliver: let a traveler search buses and hotels across every major provider, compare fairly, and book — all in one place, on any device.

The core problem: aggregation at speed

Each provider exposes a different API, with different rate limits, response shapes, and reliability guarantees. A naive fan-out to six providers on every search would be slow and fragile. We needed a unified layer that normalizes providers, caches aggressively, and degrades gracefully when an upstream is down.

Why FastAPI + Next.js

FastAPI gives us async I/O — essential when fanning out to many providers concurrently.
Elasticsearch powers sub-500ms search with relevance ranking across normalized results.
Next.js delivers a fast, SEO-friendly web app that shares types with the backend.
React Native reuses that logic across native iOS and Android apps.

The unified integration layer

Every provider is wrapped in an adapter that implements the same interface. The search service calls all adapters concurrently with a strict timeout, merges and de-duplicates results, and writes them to a short-lived Redis cache keyed by route and date.

async def search_buses(route: Route) -> list[Trip]:
    tasks = [adapter.search(route) for adapter in PROVIDERS]
    results = await asyncio.gather(*tasks, return_exceptions=True)
    trips = [t for r in results if not isinstance(r, Exception) for t in r]
    return dedupe_and_rank(trips)

Because each adapter fails independently, one slow or broken provider never takes down the whole search — it simply drops out of that result set.

Takeaways

The right architecture is the one that lets your team move quickly today and still scales tomorrow. We optimize for clarity, measurable performance, and a codebase your team can own.