The Knowledge Vault is a powerful tool designed to help create on-brand AI-generated content by storing critical company information, such as brand-related documents, data and content, and making it available to generative AI for content creation.
Unlike a traditional database that stores exact copies of images, documents and text, the Knowledge Vault is a vector database, which stores data as numerical representations, or vectors, that capture the meaning and context of information instead.
A vector database enables more intelligent searches based on the meaning behind the words, known as a semantic search. This allows AI to understand and retrieve information that is contextually relevant for the content that is being created, even if the exact words don't match.
When you upload content to the Knowledge Vault, it doesn't store the full document in its original format like a PDF file. Instead, it extracts the text from your content and converts it into vectors, positioning it in a vector space based on its semantic meaning.
This enables AI to perform more nuanced searches and generate content that is more aligned with your brand's voice and guidelines.
So, while you won't be able to retrieve the original file directly from the Knowledge Vault, if you store the external ID from a separate Digital Asset Management (DAM) system, you could, in theory, retrieve the associated file from there. The AI will still use the embedded knowledge to create content that accurately reflects your brand's identity and messaging.
The following are the key features of the Knowledge Vault:
There are three methods to add content in to the Knowledge Vault: