Nov 4, 2026
|
57 min
|
5.5k views
How do you build a Shiny app in 2027? | Dan Chen | Data Science Lab
Start from a working app made with an agent, give the agent the right context, use the modern defaults with bslib, and use version control and a simple deploy path. Learn the fundamentals so you can review the output, while treating the LLM as a tutor. Stop Building Apps Like It’s 2018!
In this Data Science Lab session, Daniel Chen provides a foundational walkthrough on creating a Shiny application using R. The demonstration begins with setting up a new repository on GitHub and cloning it to a local machine using Positron. Daniel leverages Claude to generate a basic Shiny dashboard based on a TidyTuesday dataset. The presentation covers the structural differences between the UI and server components in Shiny, explaining how user inputs are captured via the ‘input’ variable and how results are rendered back to the interface through the ‘output’ variable. Additionally, the session explores the use of ‘bslib’ for modernized UI layouts and discusses the role of AI agents in writing and commenting code, suggesting a workflow where developers use LLMs to explain complex regex or specific function logic.
The Data Science Lab is a live weekly call. Register at pos.it/dslab! Discord invites go out each week on lives calls. We’d love to have you! It’s is an open, messy space for learning and asking questions. Think of it like pair coding with a friend or two. Learn something new, and share what you know to help others grow.
Originally recorded August 25, 2026.
0:00 Welcome and intro with Daniel Chen 2:32 How Data Science Lab works (candles, golf flags, Discord) 5:14 Start a new project on GitHub 8:26 Clone the repo and keep a git folder 10:59 Git vs. GitHub 11:25 Set up Claude Code in Positron 12:14 Install Posit skills (posit-dev/skills) 15:19 Prompt Claude to build a TidyTuesday Shiny app 20:51 Skills and the context window 21:45 Run the app for the first time 22:14 Should you still learn Shiny by hand? 25:13 Run the app from the command line 27:19 Tour of app.R: packages and data 29:18 Ask Claude to explain code you don’t understand 32:24 Side questions with /btw 33:18 How a Shiny app works: UI and server 34:46 bslib page_sidebar and input controls 38:35 Value boxes, cards, and outputs 39:42 The server function 42:37 Inputs, outputs, and how they connect 47:27 Reactive expressions 48:56 Commenting code (the “why” vs. the “how”) 53:01 Commit and push to GitHub with Claude 54:22 Wrap-up and deploying with Posit Connect Cloud 56:07 Join the community, plus next week’s episode
Resources:
Download Positron (the IDE Dan is using): https://positron.posit.co/download.html Today’s repo: https://github.com/chendaniely/2026-08-25-dslab Course textbook for DSCI 532: https://ubc-mds.github.io/DSCI_532_vis-2_book/ Getting started with Shiny course in Posit Academy: https://academy.posit.co/introduction-to-shiny-r-focus Academy search page for shiny (R and Python): https://academy.posit.co/?q=shiny Scoping rules: https://shiny.posit.co/r/articles/improve/scoping/ Posit Dev Skills: https://github.com/posit-dev/skills bslib package in R: https://rstudio.github.io/bslib/ Mastering Shiny (free book!): https://mastering-shiny.org/
bslib
btw
Positron
rstudio
Shiny