Just getting rolling with ECM and trying to find the best process for batch importing a bunch of sales order related documents (pdfs, images, excel) saved in network folders.
So far I have a batch process that cleans up the folder structure so I have folders named with the sales order number and a batch import that imports those folders and all the files inside of them. That part is working fine. I am wondering if I can do more with a workflow and if there is really any benefit to it.
OCR: I have created a workflow and added the OCR action. I can see that the workflow ran but I can’t seem to search for any of the text within the documents. Not sure if this is really necessary since most of the text they would be searching for should exist in the sales order within Kinetic (customer name, part number, etc.).
Tag Order Number: Is there a way to tag the order number by pulling it from the folder name? Again is there really any benefit since it is already being stored in a folder with that order number so even if the document has a meaningless name, it can be found in the folder.
You did it the “other way” but I wanted you to know this can be done in one pass using a properly formatted CSV or XML file and the ECM client’s Batch Import tab. If you know the order number, then it would get mapped to the meta data field and can be used in a simple workflow to go grab data from Kinetic by way of a datalink. that would make the document searchable by whatever data you’ve brought back - customer, date, part number, etc.
The ECM Client OCR process isn’t a simple whole-document text indexing process. It must put the results in a multi-valued document type data field. I guess I should assume you are doing that part, so you’ll need to wait for the indexer to complete before the Search will work. We don’t do full OCR tasks, so I’m guessing a bit on this one.
Not sure - you could try some string functions on a variable set to the folder name. That should work.
Yes and no - depends on your process for locating documents. A proper folder structure means you don’t need meta data to search on. However, folder structures are not necessary for CMS’s like ECM. But… because your filenames aren’t really useful, and because you’ve not added any other useful meta data in the data fields (you didn’t mention doing it) there is no searchable index data. In that case, without a useful folder structure, then getting the order number into a field makes the data searchable.
I appreciate the feedback @MikeGross. One of our uses for ECM is to replace a document management system that we used to use. I keep finding myself trying to make ECM perform exactly the same way which is what I try to keep my users from doing.
The folder structure I am importing goes back to 2016 so the naming conventions changed many times and what is in the folder changes by the type of order so a ridiculous amount of inconsistency.
I want the old information available and then will be working on better workflows for adding the new data including meta data to try to get the added benefits.