Artificial Intelligence (AI) is having an increasing impact on content management. For several years, Nuxeo has made AI for content available through integration with public cloud offerings like Google Vision and Amazon Rekognition. We’ve seen more and more organizations realize the benefits of deploying the technology to help automate and improve functions within their organizations. The value is already becoming clear: AI can help companies simplify complex content processes while improving information management as a whole.
But as companies increasingly look to put content in motion to achieve key business outcomes, our technology continues to evolve and improve at a rapid pace.
Enter Nuxeo Insight.
This unique and more advanced offering sets an entirely new standard for AI and Content Services Platform (CSP). Now that users have more than one way to leverage AI within the Nuxeo Content Services Platform, we thought it would be worth a closer look to clearly understand how the two services differ.
Previously, Nuxeo customers had the option to integrate with existing AI providers (e.g. Google Vision, Amazon Rekognition, etc.) These services are dubbed “Generic AI.” While backed by powerful technology, these models possess no knowledge about your specific organization, the terminology you use, and how you work - because the algorithms are trained on general datasets from public sources that can be broadly applied to support almost any business use case.
We’ve all had the experience of joining a new organization and feeling like we’re in a bowl of alphabet soup for the first few weeks or months. Generic AI services are like an overwhelmed new employee that can tell if people in the meeting are happy or unhappy, but doesn’t have the context to understand why or contribute much at first.
Generic AI typically involves some form of simple, automated metadata tagging. A user can upload a document, and quickly receive basic information about what kind of document it is – a contract, for example, or a memorandum. If you upload a photo, it might apply three to four generic metadata tags to help categorize that photo – telling you it’s a picture of a car, perhaps, or a tree.
While limited, this AI still provides multiple information management benefits – it lets organizations streamline their information management processes while reducing storage, increasing security, and delivering faster and more effective searches for content and information.
As the volume and variety of content and metadata being stored by organizations continues to explode, so too does the demand for more sophisticated and detailed AI. This is another area where Nuxeo is on the leading edge of innovation.
Unlike other services in the market, Nuxeo Insight allows users to train their own custom AI models using business-specific data sets. This next generation AI service is capable of stacking a seemingly endless number of classifiers against an individual piece of content to provide significantly more in-depth, sophisticated metadata descriptions.
Suppose NBC was using Nuxeo to manage its video files of Saturday Night Live (SNL). With Generic AI, a skit with comedian Adam Sandler that originally aired in 1995 would be labeled with the exact same metadata label as video from his return to the show nearly 25 years later. The metadata only says “it’s Adam Sandler on SNL.”
Contextual AI trained with custom data sets through Nuxeo Insight, by contrast, can distinguish between all of Sandler’s SNL characters over time. The custom AI models would apply metadata informing NBC that “this is Adam Sandler debuting as Canteen Boy in 1994” vs. “Sandler’s Encore of Opera Man in 2019.”
This level of detail is much more useful. What’s more, the system is easy for customers to use – they simply upload their content, which is then used to train machine learning models. These in turn are used to tag new data as it is ingested into the system.
AI + Content Platforms = Smart Business
Infusing Content Platforms with Contextual AI makes metadata schemas much more flexible and extensible. In addition, much more metadata is being stored and used than ever before – image resolutions, language of a document, geophysical data, etc. This increased capability and the ability to utilize metadata more effectively makes the modern Content Platform highly valuable to today’s enterprise.
It’s been well established that business-specific metadata is the foundation of effective search, workflow, and other value-creation activities in content-centric business applications. The challenge has always been the manual effort and investment required to properly and accurately identify content and link it to related materials. Existing content enrichment AI services address this challenge in one of two ways. Some services are easy to deploy and provide generic metadata not based on business specific content. Others allow custom model development, but require scarce data science expertise to use.
Nuxeo Insight addresses both of these challenges head-on with trainable AI models for business users.
How will Content Platforms & AI work together in the future? What will be the best use cases for intelligent content management, and what are the major benefits it will provide? I’ll examine those questions in one of my next posts!