From events to insights
A healthcare data pipeline that carries synthetic patient events through validation, storage, and into an interactive dashboard.
Kafka / FastAPI / Prefect
The idea
An analytics dashboard is only as useful as the data feeding it. This project explores the steps between generating patient events and making them available for analysis.
How I built it
A synthetic data generator sends batch payloads through Kafka to a FastAPI backend. Great Expectations validates data in two stages, Prefect handles orchestration and retries, and Neon PostgreSQL stores the results. The services run in Docker.
Where it landed
A Streamlit dashboard for exploring patient metrics, vitals distributions, and diagnostic patterns using synthetic data.