Background: JSL-automated data analytic pipelines are great for routine data queries, transformations and visualizations. But in every organization there is a need to improvise new approaches to data analysis as well. Fully automated data analytic pipelines, though powerful, don't always help their users climb the JMP learning curve and can promote "learned helplessness"
Suggestion: I'm thinking it'd be nice to be able to use JSL to launch common JMP platforms (Summary, Fit Model, etc.) with various fields pre-populated with various columns. Although users now need to perform an extra click ("OK") to do their routine analyses, the pre-populated modal will (1) help them better understand how to use these JMP platforms for improvised analysis, and (2) streamline improvisational data analyses that utilize the routine analysis as a jumping-off point.
I can think of a few ways to hack this outcome with JSL in its current state, but those hacks are inelegant. They involve doing the routine analysis to completion and then Relaunching the platform or automating a click of the recall button.