The evolving state of enterprise content management: How AI changes the game

A recent Forrester study shows a growing number of companies feel their workers spend too much time looking for information they need – 40% today vs. 19% just five years ago. A number of issues contribute to the problem, including a highly distributed workforce, siloed technology systems, the massive growth in data, and more.

But it doesn’t have to be that way because enterprise content management systems have made great strides in that same timeframe, including with new artificial intelligence technology that makes it far easier for employees to find and make the best use of all the content the organization owns, no matter if it’s text, audio, or video.

While ECMs have always been useful, in the past, they required too much effort from users. Intuitively, it’s easy for people to understand a piece of content and classify it according to some well-understood structure. But content management tools often asked users to do things they inherently don’t like doing, such as extracting information from a piece of content and entering it into fields and tables to describe what it is. The systems worked, but not without some manual effort.

AI and related technologies, such as machine learning (ML), enable content management systems to take away much of that classification work from users. Importantly, such tools can extract relevant data even from unstructured data – including PDFs, email, and even images – and accurately classify it, making it easy to find and use. Some ECM systems have intelligent document processing (IDP) capabilities that can mimic the way an employee would read a document, extract key information, and enter it into another system for processing.

“AI enables ECM solutions to unlock valuable information from unstructured data and maximize the value of their content,” says Ericka Morimoto, Product Marketing Manager, of Hyland, an intelligent content solution provider. “Users can get business-specific answers, not generic answers like with consumer large language models, to make better-informed decisions.”

Key features of a modern ECM

Different ECM solutions often focus on varying functions or use cases, but the following are some technologies to look for in a modern ECM.

Natural language processing (NLP): As its name implies, NLP employs ML to essentially “read” a document much like your employees would. It can perform data extraction, sentiment analysis, and language detection, as well as document classification.

Deep learning for image and video technology: Much like NLP can “read,” deep learning technology enables an ECM to view images or video and identify objects, text, people, activities, and more. Want to find all the content you have that includes a photo of a particular celebrity? Deep learning can help with that.

Speech-to-text conversion: Content can take various forms, with video and audio growing in proportion. Speech-to-text uses advanced ML algorithms to transcribe audio files into readable text so it can be more easily classified by an ECM, searched, and more. For example, with speech-to-text, you can transcribe customer service calls and apply sentiment analysis to determine how agents handle sticky situations.

RESTful API for image analysis: This is another ML capability that enables an ECM to classify and label images, detect embedded objects, and extract text. An example use case would be an insurance company using it to read license plates from an auto accident photo.

With such features incorporated, ECM platforms become far more useful, especially if they work not just on-prem but in cloud environments. Additionally, search becomes easier, with users able to both search and get results in natural language, like talking to their smart phone assistant.  No matter where the content may be, the ECM solution will find it, enabling it to be put to good use.

Learn more about Hyland’s intelligent content solutions here.