Webinar replay: How carriers are leveraging large language models (LLMs) and automation to drive better decisions
Watch Now
  Everest Group IDP
             PEAK Matrix® 2022  
Indico Named as Major Contender and Star Performer in Everest Group's PEAK Matrix® for Intelligent Document Processing (IDP)
Access the Report

BLOG

How Intelligent Process Automation Addresses Unstructured Content Risks

January 15, 2020 | Business, Intelligent Process Automation, Text Data Use Case

Back to Blog

 

True story: A chief technology officer didn’t believe an intelligent automation tool could, on its own, look at a photograph and discern what type of room in a house the photo depicted. So, the salesperson conducting the demo went to the real estate site Zillow and clicked on a random listing. The CTO zeroed in on a photo showing a room that looked like a disaster area, with clothes, boxes and bags strewn all about. Undeterred, the salesperson entered the URL for the photo into the automation platform.

Seconds letter, the platform came back saying it was 99% confident the photo was of an attic space.

Bingo.

This is the power that artificial intelligence, and cognitive automation platforms in particular, can bring to business problems. The ability to understand unstructured content and apply human-like decisioning opens up new avenues for where you can apply process automation technology and removes the risk that comes from relying on humans to perform repetitive processes.

Unstructured content: What it is, why it’s a problem

Unstructured content creates problems for rule-based automation engines, including robotic process automation platforms, and OCR templating approaches, because it’s so difficult to define rules that apply to something you can’t predict. And unstructured content is nothing if not unpredictable.

By definition, unstructured content refers to content that is variable in nature. It could be contracts, Word documents, text (including emails) and images.

Take a bank statement, for example. Statements from the same bank will likely have fields that show up in the same place from one to the next – name, address, account number, start date, end date and the like. But statements from another bank are likely to be quite different. If you’re taking a rule-based approach to automating a process that involves looking at statements from numerous banks, such as a credit approval process, it’ll be next to impossible to come up with enough rules to take into account how each bank’s statement is formatted.

Another example that may be top-of-mind for most financial professionals is the imminent demise of the London Interbank Offered Rate, or LIBOR, set for the end of 2021. For many years, LIBOR has been the benchmark to set interest rates for most variable-rate loans, interest-rate swaps and other financial instruments. Companies need a way to search their contracts for language around LIBOR, understand what their risk is, and examine whether changes are required. Here again, coming up with a rule-based approach to examine myriad such contracts would be tedious to say the least, and likely impossible.

One alternative, of course, is to rely on humans to manually look at documents and images, find the relevant fields, extract desired information and enter it into another application, document or spreadsheet. As with any task involving humans, that process comes with expense and the risk of human error, whether it’s transposing numbers, misspelling a name or simply missing something important. Humans do, after all, get tired – and bored.

The opportunity in intelligent automation 

Clearly there’s a huge opportunity to be had in automating such repetitive, error-prone processes, including freeing up employees to perform more rewarding and strategic work. But it requires the ability to effectively deal with unstructured content.

As the Zillow example demonstrates, intelligent process automation offers just such a capability. Intelligent process automation incorporates technologies including optical character recognition, enabling it to “read” unstructured documents, and natural language processing, which provides the ability to analyze and understand context, much like a human does. Machine learning and deep learning technologies also come into play, enabling processes to continually learn from the past, while transfer learning enables intelligence gleaned on one process to apply to other, different processes.

It all adds up to increased efficiency for your organization. If our technology can tell an attic from a messy bedroom or basement, chances are we can help automate whatever processes you need to deal with.

To learn more, check out this Everest Group white paper, Unstructured Data Process Automation. 

[addtoany]

Increase intake capacity. Drive top line revenue growth.

[addtoany]

Unstructured Unlocked podcast

March 27, 2024 | E43

Unstructured Unlocked episode 43 with Sunil Rao, Chief Executive Officer at Tribble

podcast episode artwork
March 13, 2024 | E42

Unstructured Unlocked episode 42 with Arthur Borden, VP of Digital Business Systems & Architecture for Everest and Alex Taylor, Global Head of Emerging Technology for QBE Ventures

podcast episode artwork
February 28, 2024 | E41

Unstructured Unlocked episode 41 with Charles Morris, Chief Data Scientist for Financial Services at Microsoft

podcast episode artwork

Get started with Indico

Schedule
1-1 Demo

Resources

Blog

Gain insights from experts in automation, data, machine learning, and digital transformation.

Unstructured Unlocked

Enterprise leaders discuss how to unlock value from unstructured data.

YouTube Channel

Check out our YouTube channel to see clips from our podcast and more.
Subscribe to our blog

Get our best content on intelligent automation sent to your inbox weekly!