I have an SDE with a point feature class that has 120k locations across the entire US. I can read those points in with a normal reader, but it takes a long time ~5 minutes and I really only need to return 100 out of those 120k points.

Which 100 changes depending on the input of an excel reader. I pull in a list of 10 or 20 points from an excel sheet, take a 2 mile buffer around those points, and then spatial join that to the big list from SDE. Those 10 to 20 points could be anywhere in the US.

Is there a transformer that will let me read in my 10-20 points, then make a query to SDE/SQL with a transformer based on the location, so that I'm only asking the database for the records I need? That way I am not pulling in the rest of the 100k+ records.

Right now I am thinking I can take the lat long coordinates from the excel input and create a bounding box of sorts to query the SDE point feature class. But I don't see a transformer that can read from SDE 'inline' with the rest of the workbench.

Any thoughts would be appreciated.