Yext uses generative AI to introduce automatic content generation for its knowledge graph

Yext, Inc., the Answers Company, is adding Content Generation to its knowledge graph product, aiming to have the solution automatically and proactively generate its own content.

“Content is king in any digital experience. We’ve seen time and time again that the more content you have, the better you perform across search engines, websites, mobile apps, and other digital touchpoints. However, many teams lack the in-house resources to produce high-quality content at scale,” said Marc Ferrentino, president and chief operating officer at Yext. “We set out to solve this problem for our customers. With Content Generation, organizations can leverage the latest generative AI to dramatically increase their team’s efficiency and content output while enriching digital experiences wherever users engage with their business online.”

The company has been running a year-long pilot with several customers and plans to officially launch Content Generation in March with Yext’s Spring 2023 Release.

Content Generation uses multiple large language models including GPT-3 and existing information from a customer’s knowledge graph to automatically generate and suggest rich, business-specific content that is on-brand and aligned with writing styles or patterns found throughout an organization's content library.

Yext customers can use these features to effortlessly create descriptions, blog posts, biographies, FAQs, and more, automating what could otherwise be manual and resource-intensive content creation tasks, according to the company.

With these features, content is generated according to a specific set of inputs that are determined by the business.

For example, a healthcare system can choose to generate biographies for professionals based on known values for name, education, specialty, and work history.

Since content is generated with information from a customer’s knowledge graph, outputs are more accurate and less susceptible to the “hallucinations” that often plague general-purpose large language models, according to the vendor.

As an added layer of quality control, organizations can use Yext’s Suggestions workflow to review and approve new content as it gets created.

For more information about this news, visit www.yext.com.