Knowledge

Maintains account-level memory, allowing the agent to recall aggregate account history and context across multiple conversations.

The agent has its own authorizations across services it has access to, allowing for more granular control on the actions that the agent can take.

Implements standard RAG (Retrieval-Augmented Generation) to access company knowledge, with capabilities to scrape URLs and upload files for comprehensive information access.

Currently available as a web deployment, with upcoming support for Slack and additional communication channels (Phone, Discord etc).

You can upload documents directly to your agent’s knowledge hub. Supported file types include:

• PDFs
• DOCX (Word)

File upload limits:

• Max size: 100MB per file.
• Files are processed instantly and made available to the agent in real time.
• The agent will begin using this content immediately in relevant flows like L1 deflection, summarization, or internal queries.

Provide public URLs to web pages and documentation sites. When a URL is added, the agent will:

• Scan the primary page
• Automatically crawl and index all publicly linked sub-pages
• Continuously stream and refresh that content in the background

Once sources are added, the agent can:

• Access the full content of uploaded documents
• Read and reference all linked public pages under each URL
• Use this information to answer questions, deflect tickets, summarize threads, or inform other flows

You can view, manage, or remove knowledge sources at any time.

Giving your agent access to the right knowledge:

• Increases the accuracy and relevance of responses
• Ensures consistency with your brand and documentation
• Powers smarter AI flows with richer context

With just a few uploads or links, you turn your agent into a true expert on your business.

Knowledge fuels intelligence. And your agent gets smarter with every file and link you add.