Wrap Up

What we covered

• Shiny’s input → output → reactivity model
• Building and running your first app
• @reactive.calc for shared computations
• Dashboard layouts: sidebars, columns, cards, value boxes
• Deploying to Posit Connect Cloud
• Integrating LLMs with chatlas, shinychat, and querychat
• Testing with Playwright

Key concepts to take home

• Reactivity is automatic — Shiny tracks dependencies for you
• Start static — get the visualization right, then add inputs
• @reactive.calc — avoid redundant computation across outputs
• requirements.txt — always needed for deployment
• SQL is a great intermediate language between humans and LLMs

This workshop

Everything from today:

https://github.com/chendaniely/pycon-2026-shiny

Feedback survey

Please take 2 minutes to fill out the PyCon session feedback — it directly influences future tutorial selections.

Survey link is in your email / QR Code