Back to case study

CORDIS to Supabase Pipeline

Python, Pandas, Parquet, Supabase and data validation

Summary: A portable Python ETL pipeline that prepares cleaned CORDIS data for the live research web app and loads the final web-ready dataset into Supabase.
82,370web-ready project records
0blank programme values after validation
Portableruns outside Fabric

Objective

The pipeline was created to make the CORDIS project independent from temporary cloud trials. It allows the cleaned web dataset to be rebuilt locally and loaded into Supabase whenever needed.

Pipeline Flow

Outcome

The pipeline supports the live CORDIS Research Explorer app and gives a repeatable path to rebuild the dataset without depending on a Fabric trial account.