Opening the door to truly intelligent content automation

In the same way you might ask a ChatGPT Plus-based voice assistant for information, knowledge workers are asking business questions out loud, rather than typing them or framing queries, and getting useful replies.

Clinicians will ask their patients’ clinical reports to discuss and organise themselves in priority order based on patients most at risk of cancer. Lawyers writing a contract will ask their legal documents if any of them contradict or conflict with this new contract. And insurance claims clerks will be asking a file that has arrived in their inbox what it is about, what policy it relates to, and whether the request is reasonable.

The file might even be in a foreign language, but your questions and the feedback to you is in your local dialect, natural language or accent.

What I’m describing here is what it’s like to live and work in a world of truly intelligent documents.

These are self-aware, communicative documents whose insights and content are so interrelated and self-aware that they are effectively “conscious” and able to share information about themselves.

Instead of searching, we are able to ask business questions of billions of intelligent documents and other items of organisational content and get a summarised intelligent response.

And to a large extent, this is already here. Customers of ours already have automated systems to read all their incoming emails, categorise and understand embedded documents and intelligently act on them. For example, global manufacturer SEW-Eurodrive, which manages hundreds of millions of documents, is able to understand custom orders and use them to automate order workflows, including automating the creation of a CAD design, and the scheduling of manufacturing and delivery. And for organisations that have beyond, say, 60 billion documents to manage, intelligent documents can help them a lot!

I call this the dawning of the age of the ‘sentient document’ – a revolution that could be the most important advance in the automation of documents since the invention of the printing press.

The drivers of change are clear enough: deep learning will, at an increasingly faster pace, enable the automated understanding of documents. Large language models and generative AI are helping to enrich the two-way dialogue between human and computer. Combining this innovation with intelligent document management platforms, we can now communicate with our corporate repository in natural language to query, search and automate workflows around documents.

Sentience will be exhibited by general business content (purchasing orders, invoices, job applications, change of address forms, customer onboarding forms, etc.), but also by specialist content in fields like medicine or law. And the advent of the sentient document couldn’t be more timely. By 2025, global data creation is projected to grow to more than 180 zettabytes. And note that one zettabyte equates to a trillion gigabytes.

Faced with this deluge of information, having advanced AI assistance will prove invaluable in helping to navigate, process, and derive insights from this data.

But this shift to super-intelligent AI assistance at the scale and pace required to manage the projected data growth will not just happen naturally. Content owners will have to facilitate this paradigm shift. Organisations must also exercise caution to be sure that sentient documents won’t increasingly subsume vital work intended for a human intellect.

However, it’s not just AI that is needed here to make sentient documents come to life. Organisations need a platform with AI at the core that combines highly scalable content management with end-to-end process automation. We call this platform approach ‘Intelligent Content Automation’.

There are 5 key requirements for bringing Intelligent Content Automation (ICA) to life in today’s enterprises:

1. You need to fuse AI, automation & content – Traditional enterprise content management platforms cannot deliver the combined AI, automation and content management capabilities. An Intelligent Content Automation platform harnesses these capabilities and unifies content based on metadata.
2. Metadata is wisdom – Metadata is information about documents. In an ICA platform, metadata can be generated, stored and managed as AI decodes and understands documents. It can be consumed by generative AI to create intelligent output. Metadata can fuel workflows and intelligent searches. AI is the fundamental building block of wisdom for intelligent documents.
3. A backbone of wisdom to unlock silos – The modern enterprise is driven by diverse best-of-breed applications used across the business. Examples include SAP as the ERP in the Finance department; Salesforce as the CRM in Sales; Workday in HR or Microsoft Teams for collaboration. Problems emerge when we want to gain enterprise-wide wisdom across these apps. For instance, can we determine a customer’s upselling potential based on invoice data? Can we use this data to assess the performance of Sales? The answers lie locked away in disconnected document and process silos. To build a backbone of wisdom for the enterprise, an ICA platform connects all these systems to deliver contextualised insights.
4. Lighting up dark data – Dark data is data that resides in the business, but you’re not able to use or link to it. Consider all the Teams calls that take place, for instance, with great ideas, collaborations and documents. Once finished, these likely lie dormant and forgotten, like closed boxes in the cloud. Some estimates are that 54% of business content on computers is dark, which contributes to stats like the $3.1 trillion of lost revenue from poor data quality in the U.S. alone. Content platforms must connect to and mine these dark documents, extracting their metadata and combining it into our wisdom backbone.
5. Working with billions of documents – Finally, enterprises are now typically accumulating billions, not millions, of documents. We need a platform that can do all of the above at a massive scale and can support streaming document intelligence.

Sentient documents are the killer app not just for the enterprise content management sector, but for business globally.

But to get there, we need to deploy next-generation intelligent content automation platforms, as that’s the best way to add intelligence to communication with documents, automate document processes and make document discovery as easy as asking your voice assistant about the weather.