How Anchorage Built Alaska’s Vaccine Finder with R
summary
A one‑week prototype becomes a statewide vaccine finder built with R that scrapes appointment pages and serves a simple mobile HTML table, connecting tens of thousands of Alaskans to appointments and handling ~1,000,000 pageviews with peaks of 20–40,000 pages/day. Core tools include rvest & httr for scraping, dplyr, stringr, lubridate for data wrangling and a Heroku R buildpack plus S3 to run jobs 24/7. About 5 million HTML pages are scraped, sources are modularized into six inputs, and error handling keeps the site running.
AI-written summary · about
chapters (22)
- 0:00Alaska vaccine rollout challenges
- 1:32Anchorage Innovation Team
- 2:15Senior citizens access problem
- 2:58Weekend prototype in R
- 4:01Simple HTML frontend
- 4:43Prototype testing with call center
- 5:24Unexpected launch expansion
- 6:06Statewide vaccine finder launch
- 7:12Governor's announcement handling
- 7:54Success through adaptability
- 8:37Traffic spikes and timelines
- 9:19Million page views scaling
- 9:40Scraper to stable system
- 10:22Data pipeline modularization
- 11:0424/7 appointment tracking
- 11:47Eligibility simplification
- 12:49MVPs and prototypes power
- 13:33Unexpected data set value
- 14:15R for rapid prototyping
- 15:23Speed and flexibility matter
- 15:46Build fast appreciation
- 16:06R as a superpower
related talks
16mHow Anchorage Built Alaska’s Vaccine Finder with R
51mViteMaDose: the missing link in the vaccination campaign
01h 46mBrian & Brock's App Adventure
30mHow to Build Agentic Web Research Assistants | Live Demo & Code
18mBuilding a Real-Time COVID-19 Surveillance System with R
25mLow-code development with APEX
43mLet's Build! Patient Mediated Data Connectivity
20mUsing Python and Gemini to Visualize Health Alerts on an Accessible Website