Airbnb Data Pipeline.

01

Context & problem

This project processes public Inside Airbnb data for Dublin. The goal is to turn raw listing and review data into a structured foundation for analysis, rather than analyse directly from disconnected CSV files.

02

My role

I built the end-to-end ELT workflow, data model, automated checks, and analytics delivery. The project brings together ingestion, warehouse transformation, orchestration, and dashboard reporting.

03

How it works

Python loads source data through Google Cloud Storage into BigQuery. dbt transforms warehouse data into an analytical star schema.

Airflow orchestrates the workflow and Docker packages the environment. Looker Studio presents market, location, and listing information from the resulting analytical data.

04

Key decisions

I used layered transformations and a star schema to organise the warehouse for analysis.

Automated dbt checks cover nulls, uniqueness, accepted values, price validity, and referential integrity, placing validation within the data workflow.

05

What I built

Processed more than 6,900 listings and 300,000 reviews in the project dataset. These figures describe the processed dataset, not a live service.

Delivered an automated pipeline, a warehouse model, quality checks, and a dashboard for exploring market and listing information.

06

Limits & next steps

The source is a public dataset rather than a live booking feed. Dashboard outputs reflect the source snapshot and its limitations.

The repository identifies incremental models and infrastructure automation as future improvements. This case study does not claim measured business impact from the dashboard.