Developing a Prototyping Competency in a Statistical Science Organization
summary
Argues that a deliberate prototyping competency speeds adoption and reduces risk by proving ideas fast; a two-person, scrappy team runs 2–3 projects at once and targets 10 prototypes per quarter. Uses Shiny plus an async crew worker pattern to offload long-running jobs, builds a POC that is open-sourced and extended, and moves most work only to alpha/beta readiness before handing off. Uses an organizational hierarchy of needs, OKRs, fear-setting, and quarterly scorecards to measure and iterate.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:57Building prototyping capability
- 3:47Statistical hierarchy of needs
- 6:17Prototyping process
- 8:49Prototype finish line
- 9:35Prototype maturity scorecard
- 11:03Small team operations
- 12:29Risk management
- 15:05Key prototyping lessons
- 19:26Q&A
related talks
32mPanel: Make The Most Of Data Science
25m"Data Science & Agile: Strategies for Working Together"
23mData science as a team sport
01h 03mExperimentation in Data Science
49mHow to get smarter while doing your job
20mWhat I learned from running Shiny in production
38mStarting a Data Science Project! How Do We Begin?
25mReproducibility Around a Web Application