Skip to main content

Knowledge Collections

A knowledge collection is a named group of documents that Adam - Menta's AI mentor - can draw on when working with your learners. When a course needs to look something up - to give feedback, answer a question, or ground a simulated conversation - Adam uses only the knowledge collections linked to that course, and nothing else. Knowledge collections are how you put your own material in front of Adam, so his responses and feedback reflect your company's content rather than generic knowledge.

You might create one knowledge collection for your product documentation, another for your sales methodology, and another for your objection‑handling guides. Each is a self‑contained knowledge base that you fill with files.

This chapter covers:

  • Managing collections - Browsing, searching, creating, and editing knowledge collections.
  • Uploading files - Adding documents to a collection, the formats we accept, and the requirements every file must meet.
  • Best practices - How to structure and prepare your documents so Adam gets the most out of them.

Where to find collections​

Open the Knowledge Collections entry in the sidebar under the Resources group. You'll land on the collection list, which is the starting point for everything in this chapter.

How collections relate to courses​

A knowledge collection is linked to a course, and a single course can draw on several knowledge collections at once.

You choose which course a knowledge collection belongs to when you create it - see Creating a collection. For a course to use this knowledge, at least one knowledge collection has to be linked to it.

What happens to your files​

Before a file is added to a knowledge collection, Menta automatically checks it - scanning for viruses and malware and confirming the file isn't damaged. A file that fails these checks is rejected and never reaches the collection. Files that pass are processed so that Adam can use their content. You don't need to do anything while this happens - each file shows its current status as it goes (see What happens after you upload).