—
EUR
Waiting for first observation
EVENT-DRIVEN FLIGHT MONITORING
Flight Price Intelligence
An event-driven data engineering platform that continuously monitors flight prices, publishes immutable events through Redpanda, stores historical observations in PostgreSQL, and exposes analytical insights through FastAPI.
Designed to demonstrate scalable streaming architecture rather than simply displaying airline prices.
WHY I BUILT THIS
Building a Production-Style Streaming Pipeline
This project demonstrates how a modern event-driven data engineering architecture can monitor airline prices automatically instead of relying on manual searches.
Problem
Monitoring multiple travel windows manually quickly becomes repetitive and inefficient.
Solution
Every scheduled search becomes an immutable event streamed through Redpanda and stored inside PostgreSQL.
Goal
Build a scalable streaming architecture capable of monitoring hundreds of routes and generating future fare-drop alerts.
Loading monitoring coverage
Waiting for pipeline event
Lowest stored fare across all windows
MULTI-WINDOW MONITORING
Tracked travel windows
Select a travel window to inspect its historical fare movement and current itinerary.
Loading travel windows...
SELECTED WINDOW
Price history
Select a window to inspect its history.
Window low
—
More observations are required to display
a price trend.
EVENT LOG
Recent observations
REDPANDA STREAM
| Observed | Travel window | Route | Price | Airline | Kafka offset |
|---|---|---|---|---|---|
| Loading observations... | |||||
SYSTEM DESIGN
From scheduled search to live insight
Each scheduled query becomes an immutable event, passes through a Kafka-compatible broker and is independently consumed into analytical storage.
01
SerpAPI
Flight search source
→
02
Python Producer
Event creation
→
03
Redpanda
Event streaming
→
04
Python Consumer
Validation and loading
→
05
PostgreSQL
Historical storage
→
06
FastAPI
Analytics delivery