Future-Proofing Your KM Initiatives
After the interactions are resolved, the information gets fed back into the system. “Agent Helper looks at a variety of factors including timelines, discussions of past cases, and customer sentiment,” continued Arora, “and creates articles that flow back to the customer.” Similarly, if a Slack channel continues to surface the same question, that thread can be converted into an article and distributed to the group automatically.
Both cases point to the same underlying change—content creation is becoming a byproduct of usage, which aids in future-proofing because it does not rely on a team keeping up manually. The same applies to search analytics. “Search isn’t just a retrieval layer, it’s a diagnostic signal,” added Dhar. For example, a result that surfaces often but is rated as not helpful is a signal that the underlying content needs work. “If there are searches that produce no results, then content gaps can be identified. The feedback loop and content updating should work together.”
AI will be creating more of the content over time, at a scale that makes full human review impossible. “There needs to be a confidence level and risk scoring to determine what goes into the human pipeline and what can be safely delegated to an agentic review,” noted Dhar. At the same time, more information is also being consumed by AI agents, so content has to be more precisely structured than when search presented humans with a list of documents and the human had to find the answer. “In order for AI to produce the correct answer, every knowledge chunk needs to have its own metadata,” Dhar stated. “A query should be directed to a specific answer and not have another answer be competing [with it].”
Because many organizations do not have the resources to prepare content at this level of detail, SearchUnify has launched a knowledge cleaning and enriching service. “Our customers provide us with their knowledgebase,” Dhar explained. “We ask for their content guidelines and enforce certain guidelines of our own, and then we give them back their knowledgebase with a better content quality.” Along the way, SearchUnify marks the articles that have broken links and can also identify new knowledge creation opportunities.
The Human Element
Even as some workers are concerned about losing their jobs to AI, others worry that they don’t have the skills to use AI in ways that their existing jobs require. In a study conducted by Resume-Now, an AI resume builder, 43% of survey respondents felt that their current AI skill level was behind what they needed to stay competitive in their jobs. Nearly as many (39%) worried that their co-workers who adopted AI skills faster than they did would have a competitive advantage (resume-now.com/job-resources/careers/ ai-skills-race).
“One of the strongest results was that employees crave clarity on what skills are needed,” said Keith Spencer, a CPRW (certified professional resume writer) career expert with Resume-Now. “There is a lot of AI anxiety in the workforce, but people don’t know which skills to prioritize in order to retain job security.” He noted that many employees would benefit from structured learning settings in which they could build their skills.