From Data Science Notebook to Shipped Feature (DataLoom FYP)

The Gap

Many data science projects stop at “nice Jupyter notebook”. For DataLoom, we wanted:

• Real users
• A real UI
• Repeatable pipelines

Steps We Took

1. Stabilized data ingestion (ETL)
2. Defined a minimal set of useful charts
3. Wrapped models behind a simple API
4. Built a Streamlit UI that non-technical users could understand

The result: an A+ FYP that actually felt like a product.