I've been researching this for a very long time, and figured I'd bounce this off the community in case anyone has a lot of knowledge in this department.
My company has to handle a lot of paper invoices. We work with waste companies (Waste Management, Republic, Allied, etc), and get an absolute ton of invoices per month. We then have to turn these all into data in our systems to make it through AP/AR, and we've been doing this by manual data entry.
I've been looking at various OCR web services, the best of which has been Captricity. With Captricity, I can upload a blank template of an invoice (so lets say Waste Management), and then month to month, I upload the invoices, and it pulls data from pre-mapped sections of the page, and outputs to .csv
That's all great, but my problem is in template management. Think of it this way - you eat at your 5 favorite restaurants, but the format of your final bill could be 500 different versions; did you get a steak? No appetizer? Pepsi instead of Coke? 4 guests instead of your usual 2? Oh, and did you eat at Applebees instead of Chilli's, because that changes the whole receipt layout...
Waste bills are the same way, and very soon I could literally have 1000 templates for our business. Theres no way I can ask an employee to scroll through 1000 templates for each invoice and match it- they may as well do it by hand!
So the question is, is there any web application or service to read (OCR) the Invoices (different variations) and export them into the data (e.g. csv file)?