I am becoming increasingly enamored with the Technical Difficulties podcast by Gabe Weatherhead and Erik Hess.
They do a great job tackling a single topic every episode.
A recent show is about a topic that is, of course, near and dear to my heart: Scanning.[1]
During the episode, Gabe described his scanning workflow, and how he decides what goes into Evernote vs. what gets saved to an encrypted disk image:
Gabe: You put in a piece of paper and hit the scan button. For me, most everything goes to Evernote unless it has sensitive information like Social Security Numbers, Bank Routing Numbers. Those I don’t put into Evernote.
Erik: OK, so you’re totally away from your computer. You walk up to your pedestal of scanning and you just put the paper in and you hit the button. And… how does it know “Oh, there are Social Security Numbers on this so I shouldn’t dump it into Evernote.”?
Gabe: No, it still requires a human. We don’t live that far in the future. It still requires me to decide this is sensitive so this goes into the stack next to the scanner that I’ll process manually later…
This got me thinking: why can’t we do something like that? On the Mac, a tool like Hazel should be perfect for this, so I whipped up a test workflow. Something like this might be possible on Windows with File Juggler, but I haven’t tried it.
First, Make a List
You’ll want to take some time to think about what you don’t want to go to Evernote. Take a look at your bills and other sensitive documents. What numbers (for example, SSNs, SINs, credit card numbers, bank account numbers) do you want to filter out, and how do they show up on your documents?
Next, Set Up Hazel
I set up a Hazel workflow that watches for documents in my scan inbox folder and examines them – if any of those documents do not contain the text from the list I made in the first step, it sends them to Evernote and deletes them.
If the documents do contain that sensitive information, it just leaves them alone for your future processing.[2]
The second condition is a nested condition. To create one of those, hold down the Option key when you go to hit the Plus icon beside the “Kind Is PDF” line.
Inside that nested condition, you will replace my dummy numbers with the text you came up with in your “sensitive stuff” list.
In the Do the following… section, it will run an Applescript to add the document to Evernote.
Here is a screenshot of that step when I hit Edit Script:
Here is that code for your copying-and-pasting pleasure.
tell application id "com.evernote.evernote"
activate
create note from file theFile
end tell
Limitations
There are some limitations to this approach. The first: for all this to work, your documents need to be searchable before processing. If you do your OCR later (which I believe Gabe does), this probably will not help.
The second: you need to make sure your list in the Hazel rule is accurate. Otherwise things may go to Evernote that you didn’t intend. The easiest way to do this is to check your OCRed text when building the rule. Here is the way I do it.
Do you process sensitive documents differently than non-sensitive? What do you do?
(Photo by VH Hammer)
-
By the way. Whenever I listen to a podcast about going paperless, scanning, Evernote, and the ScanSnap, I end up yelling at my iPhone about the things the hosts are getting wrong. This is the first podcast I can remember that nailed everything. ↩
-
If you don’t know how to work this Hazel thing, I hear there is a pretty good webinar about going paperless with Hazel. ↩