The Next Generation of Knowledge Management
Low-Code Application Building
Implicit to almost every significant use case relying on advanced machine learning models is the need to centralize, or consolidate, knowledge from respective sources. Whether that process involves a knowledge graph, vector store, or some other infrastructure, “GenAI can help a lot with knowledge management if your GenAI can talk to every database you have,” Aasman said. “But, GenAI needs to know what is the schema of every database and how to transform the data so I can combine it with data from another database.” When organizations adopt approaches or solutions that can accomplish these objectives, they can apply their knowledge much more meaningfully.
Lake described a process in which users compile structured and unstructured data from videos, images, technical documentation, best practices, emails, documents, and other sources for a business function and “take this huge mass of documentation into a rendering of, ‘Here’s how we would define your processing in stages and steps. Here’s how we would define your data model based on all this information you ingested.’” There are even systems that can take this information, the data model, and the stages of a process Lake described, and create, in near real time, an application from it, like that for insurance claims processing.
The Present Trajectory
There’s no denying that the next generation of KM will involve a copious number of generative models. However, it will do so within the parameters of the underlying goals of the KM profession. More importantly, perhaps, it will do so when properly aligned with other infrastructure, including knowledge graphs, vector databases, document management systems, data lakes, and more. Although this composite of infrastructure and technologies is not quite at the point in which it can fully automate the entirety of a valued business function, it can offer a more than credible launching point for an ever widening range of applications.
Smith likened the next generation of KM to coaching. “We’re going to end up with what’s kind of like coaches, where you ask a question and you get coaching. I’m trying not to use the word ‘answers’; I’ll say a starting point. As a knowledge worker, I’m tapping my system and it’s giving me a suggestion saying, ‘I think you should do this,’ or, ‘This is what other people in the organization do.’”
Whether or not knowledge managers and KM technologies will evolve into coaches, the essence of KM remains not in the tools and technologies themselves, but in how their usage advances the knowledge storage and access needs of the organization.