Create and manage Knowledge Bases

Last updated: September 2, 2026

Add and manage Knowledge Bases from the Context Manager section of the platform.


Create a Knowledge Base

To add a new Knowledge Base:

  1. In Profound platform, go to the Context Manager section in the left navigation bar. Select Knowledge Base tab at the top of the screen.

  1. Select + Add Knowledge Base.

  2. Enter a name for the Knowledge Base.

  3. Select Create to confirm.

Once created, upload content and organize your files.


Add files to your Knowledge Base

There are multiple ways to upload content to a Knowledge Base. Once uploaded, materials are indexed and made available to every part of Profound.

Upload files

To add a file to your Knowledge Base:

  1. Select an existing Knowledge Base.

  2. Select + Add Files.

  3. Select the type of upload.

The available upload types include a URL scrape, a direct file upload, a website crawl, a file import from an external source, and pasting text into an editor.

Scrape Specific URLs

Paste individual URLs to extract and store content from live webpages. This method is useful for importing content from sources such as blog posts or public product pages.

To scrape a single page:

  1. Select Scrape Page.

  2. Enter the full URL of the page you want to import.

  3. Select the sync cadence: how often do you want Profound to refresh the information from this source.

  4. Click Import URL to begin the process.

This option only scrapes the specific page referenced by the URL. It does not scrape every page within the domain. Use the Crawl Website option to scrape a full site.

Upload files directly

Upload files directly into your Knowledge Base. This method works well for documentation such as guides, policies, or internal reference materials.

To upload a file:

  1. Select Upload Files.

  2. Use Browse Files to select a file from your device, or drag and drop a file into the upload area.

Crawl Website

Index an entire site using a sitemap, or by crawling pages across a domain. Best used for documentation sites, help centers, or product catalogs.

To crawl a website:

  1. Select Crawl Website.

  2. Choose one of the following options:

  • Sitemap – Points to a sitemap XML file, allowing the system to import all pages listed in the sitemap.

  • Crawl – Progressively indexes pages across a domain. When selecting this option, configure additional options:

    • maximum number of pages to crawl, the default is 100

    • maximum crawl depth: the number of levels of the site structure to crawl (1 = crawl just the root page, 2 = root page and other pages linked from it, and so on). The default crawl depth is 3

Pages must be publicly accessible and not protected by login, authorization, or other credentials to be crawled.

Crawling pages rendered with JavaScript has some known limitations. Some sites (usually belonging to large enterprise companies) block crawlers. If this is your case, consider adding knowledge in other ways.

Import from an external source (Notion, Google Drive)

Integrate your workspace, select files in the file picker, and import them to your Knowledge Base. Best used for internal documentation, similar use cases to direct file upload.

To import files from an external source:

  1. Select Import from Notion or Import from Google Drive.

  2. If prompted, sign in to authorize the integration.

  3. Select files in the file picker.

  4. Click Import.

Paste text

Manually create a document by pasting or entering text directly in the text editor. Both structured and unstructured text are supported. Best used for creating new documents from scratch, or as a workaround when other methods don't work.

To paste text as a document:

  1. Select Paste Text.

  2. Provide the following information:

    • Document name

    • Document format (Markdown or Plain Text)

    • Content

  3. Select Add Document to save the file to the Knowledge Base.


Manage files and content

Knowledge Bases follow a folder structure modeled after a standard filesystem hierarchy. This lets you organize files the same way you would in a typical operating system, by creating folders and subfolders.

Within each Knowledge Base, you can:

  • View indexed files

  • Add tags to files

  • Create new files

  • Delete outdated material

  • Download previously uploaded content

This structure makes it easy to organize documentation by product line, brand, or other logical groupings. Well-organized content improves search relevance, enhances Agent performance, and supports more effective long-term knowledge management.