Early in my career, when I was a freelance writer and market researcher, I spent a sizable chunk of my work week sending and receiving paperwork that wasn’t my actual work – invoices, tax forms, and so forth. It was the bane of my existence.
I grew up photocopying or filling out and signing forms in hard copy, then mailing them back and forth or handing them to a clerk at a counter. So when electronic forms came along, they were a big boost in efficiency. Type some words on a pre-populated form, hit a few keys to send it, and you’re done!
The problem was when the two worlds overlapped – when a form needed an ink signature or an image from some other document (let’s say, receipts to get reimbursed for taxi costs). Those cases cost me so much time downloading, printing, signing, stapling, copying, scanning, and uploading. Often I had to re-submit forms multiple times for a single process, if clients couldn’t read my handwriting or the scan wasn’t in a format their computers could handle. I felt terribly put-upon.
Then I realized: however tedious it was for me to spend an hour or two fussing over these forms, wasn’t it worse for the clerk who had to process them all day?
From wet ink to electrons
What I described above is a weak link in otherwise automated business processes. Employees spend entire workdays keying in data from a box on one form or spreadsheet into a slot on another, or validating ink signatures. This problem crops up across every industry, whether you’re talking about paying invoices, onboarding employees, filling orders, or balancing the books. (By some figures, having an office worker in the U.S. key in data costs about 63 cents per document. That adds up fast: the U.S. manufacturing industry spends more than $1 billion a year just on data entry.)
The wet ink-to-electrons gap hits particularly hard in some paper-intensive industries such as health care (e.g., coding medical records), insurance, government (where many public records are still in paper or microfilm format), and banking.
The first attempt to bridge this gap came in the 1990s, when optical character recognition (OCR) became common. OCR scans and records information off hard copies and interprets those tiny squiggles as meaningful letters, numbers, and symbols that add up to usable information.
This basic content capture eliminates a lot of manual data entry, along with the mistyping and other errors inherent in the process. But it can also add errors if it misinterprets what it “sees,” or doesn’t deliver its content in formats that users can easily work with.
The next step in the evolution of content capture technology is intelligent capture, which aims to cut the error rate, often through multiple data extraction engines that improve OCR accuracy and rendering. Intelligent capture further adds workflow functions to route documents based on their type or keywords they contain for validation to the appropriate users. This approach dramatically streamlines document management processes that cost businesses time and money.
However, even intelligent capture solutions face limits when they’re confronted with complex or ambiguous workflows that don’t offer easily defined keywords and classifications they can “read” during capture. Figuring out who gets a document next, and what they should do with it, often requires being able to read and evaluate the overall content so that the item (complaint form, records request, complicated invoice, etc.) can be handled appropriately.
That’s where artificial intelligence (AI) comes in.
Adding judgment to content capture
OpenText™ Captiva™ is the market-leading intelligent content capture solution. Its smart features include highly accurate character rendering, the ability to read text off wrinkled, torn, or smudged originals, and automatic highlighting of key terms such as customer names and policy numbers.
Now Captiva is teaming up with the OpenText™ Magellan™ AI-powered analytics platform to offer a smart capture solution that’s enriched with AI and powerful analytics to offer a better end-to-end workflow, from content ingestion to classification, routing documents to the appropriate back-end systems, spotting exceptions, validating edge cases, and creating action items. This AI-Augmented Capture solution can produce system data and metadata for dashboards and reports, as well as contextual data for Magellan’s text mining engine, all in real time.
AI-Augmented Capture from OpenText, which combines Captiva’s powerful, flexible content capture functions and Magellan’s AI-augmented text analytics, can offer instant, intelligent document classification and routing of millions of pages a day. Routine items can be sent on their way faster, without taking up an employee’s time.
For a guided tour through OpenText AI-Augmented Content Capture, try our 20-minute webinar on demand. It covers best practices for cognitive capture of electronic and hard-copy documents. You’ll see live examples of the solution in action. And you’ll see how simple smart content capture can be.
