Daniel Perelman

About

I am a researcher in the PROSE team (opens in new tab). I got my Ph. D. from the University of Washington in 2015, advised by Dan Grossman and Sumit Gulwani.

My current work is on automatically inferring structure from visual layouts. This work has been surfaced in the PDF connector (opens in new tab) in Power Query, and I continue to work on related domains. Previously, I worked on Transformation.Text, the PROSE feature for learning string transformation programs by example. I added support for semantic (datetime and number) transformations to extend it past the capabilities of the FlashFill feature in Excel it was originally based on.