Data Science Dashboard/Web App Alternatives

Whenever I need to build a dashboard, I have always relied on a mix of Python-based libraries (e.g. Dash, Streamlit) and commercial software (e.g. Tableau, PowerBI).

This weekend, I set out to see if there are other alternative solutions that I could discover if I were to do some explorations.

If you only need a dashboard, a promising solution is to use Apache Superset. It is open source, allows no-code dashboard creation and has a comprehensive library of supported charts.

If you need a full-fledged interactive web app, I found using Vite with Shadcn to build a React web app quite a pleasant experience.

For someone who has only worked with Twitter Bootstrap for web frontend like me, Shadcn is like a super-charged React-backed equivalent for me.

But regardless of which solution you used, the biggest speed bottleneck for a dashboard is typically the database setup, which requires careful optimisations based on your needs.

Changing the dashboard solution is not going to magically solve any slowness in a dashboard.