How I ended up telling a room of knowledge management professionals that they should give up on knowledge management as they know it.
I recently found myself at KM AI, a conference for knowledge management professionals. Just as I was about to take the stage for what I had been told would be a panel discussion, I was informed that the format of the panel was in fact nothing like a panel. Instead, I would be giving a solo presentation with slides—five minutes before showtime. It felt like I was about to audition for an improv show and someone had suggested the topic knowledge management. How fun.
As I prepared for this talk, I considered crafting something polished and diplomatic. Instead, I decided to share what I believed this audience truly needs to hear: traditional knowledge management (KM) approaches are fundamentally limited, and simply adding AI to these existing frameworks won't solve our core challenges.
This perspective may challenge those who have invested years building what we've come to call a "single source of truth." Throughout this conference, many speakers emphasized the importance of such centralized knowledge repositories, while KM vendors enthusiastically promoted their AI-enhanced solutions as productivity breakthroughs.
Frankly, they’re wrong.
The "single source of truth" paradigm fails to acknowledge how information actually exists in modern organizations. Knowledge doesn't neatly reside in SharePoint or EGain—it flows dynamically through Salesforce, Slack, Teams, SAP, Oracle, Workday, and countless other systems. It also leaves the organization with employees who depart. Our information ecosystem is inherently distributed and constantly evolving.
What organizations truly need isn't to augment individual systems with siloed AI capabilities. Rather, we need an “organizational memory” that is able to touch our entire information landscape. This approach allows employees to discover and utilize knowledge regardless of its location, eliminating the cognitive burden of navigating multiple systems.
The future of knowledge management isn't about building better containers—it's about creating smarter connections.
Now, don't get me wrong. Traditional knowledge management systems aren't entirely obsolete. We use Confluence at Yurts. They're excellent for certain things, like:
These systems shine when users know exactly what they're looking for and where to find it. They're the librarians of the digital world—great at organizing and cataloging, but not so great at connecting the dots between disparate pieces of information.
The problem arises when we expect these systems to be all-knowing oracles, capable of understanding context, interpreting nuance, and pulling relevant information from across the entire organizational ecosystem. That's like expecting your local librarian to not only know where every book in the library is, but also to recite passages from your personal diary and your colleague's email from last week. It's a tall order, even for the most well-organized knowledge base.
At Yurts, we're tackling this challenge by developing AI platforms that unify data across all systems. This approach recognizes that information lives everywhere and that effective knowledge management requires integrating these diverse sources. We're not trying to be "thought leaders" here—a term that usually makes me cringe and think of Steve Jobs wannabes. Instead, we're pragmatists who recognize the messy reality of organizational knowledge.
Our approach is more akin to creating a really smart, really fast research assistant who has access to every system in your organization. This AI doesn't try to force all information into one place; instead, it knows where to look for what you need, can understand the context of your request, and can pull relevant information from multiple sources in real-time.
During my presentation, I highlighted three essential qualities for any AI solution to succeed in this complex environment. Yes, Yurts does them all - but that wasn’t the point. If you’re going to go the AI route with any vendor, whether its Yurts or someone else, make sure you that solution has these things:
Let's consider a real-world scenario to illustrate the difference between traditional knowledge management and the Yurts approach:
Imagine a customer service representative trying to resolve a complex issue. With a traditional system, they might need to:
This process is time-consuming, frustrating, and prone to errors. The rep might miss crucial information simply because they didn't know where to look.
With a Yurts AI Assistant, the rep could simply describe the issue to the AI assistant. The AI would then:
This approach not only saves time but also ensures that decisions are made based on the most comprehensive, and up-to-date information available across the entire organization which improves customer experience and uplevels reps throughout your organization.
Whether you choose Yurts or another solution, it's crucial to consider these factors. The future of knowledge management isn't about forcing all data into one place; it's about creating AI that navigates the complex landscape of information we encounter daily.
Most knowledge management companies will view this approach as an existential threat, but that's a shortsighted perspective. Horizontal AI integration doesn't threaten their existence—it elevates their relevance. Forward-thinking KM providers (including one we’re currently talking to about a partnership) recognize that enabling seamless cross-system connectivity represents the evolution of their industry, not its replacement. Those embracing this integration will ultimately emerge as market leaders, while those clinging to isolated environments risk obsolescence in an increasingly connected enterprise ecosystem.
And as for that unexpected presentation? Sometimes, being thrown into the deep end can lead to unexpected insights. The key is to adapt, process, and deliver something meaningful. In the end, that's what matters most—not how many buzzwords you can use or how "thought leader-y" you can make yourself sound.
As we move forward in this rapidly evolving field, let's focus on creating solutions that work in the real world, not just in carefully curated demos or TED Talk-style presentations. Let's embrace the chaos of information, harness the power of AI to make sense of it all, and build tools that truly support our work.
After all, in the world of knowledge management, it's not about looking like you have all the answers—it's about having the tools to find them, wherever they may be hiding.