Plotly in production (Pokémon Dashboard)

• Interactive Python + Plotly/Dash dashboard (Pokémon) shipped without writing JavaScript.
• Focused on real trade-offs: performance, UX, and maintainable structure.
• Delivered a reproducible repo/template with patterns like pre-aggregation and layout reuse.

• Layered code structure: load → compute → figures (testable and readable).
• Pre-aggregation + layout reuse for performance.
• Consistent UX: empty states, labels, and sensible defaults.
• Separation of Plotly Express (speed) vs graph_objects (fine control).

Many teams want web interactivity without a full frontend stack. Plotly (and Dash) gets you there quickly in Python.

This case summarizes real patterns and trade-offs using a dashboard as an end-to-end product: data, figures, UI, and deployment.

• Layered architecture: load → compute → make_figures to keep code readable and testable.
• Compact controls and views: enough to explore without turning into an “infinite dashboard”.
• Performance: pre-aggregation and layout reuse to avoid recomputing everything on every interaction.

• Plotly Express for fast iteration; graph_objects when you need fine control.
• Export/share as an artifact (HTML or repo) to make distribution frictionless.

• A reproducible repo that doubles as a template for future dashboards.
• A clear checklist of limits (large data, extreme customization) and mitigations.
• Transferable practices: visual consistency, empty states, and basic accessibility.

• GitHub repository