1 comment

[ 3.1 ms ] story [ 15.8 ms ] thread
How does this work under the hood? Is it OCR plus semantic mapping, or something else?