Connected Knowledge is in beta. Behaviour may change as we extend supported integrations and scale performance.
What is Connected Knowledge?
Knowledge connection is a fast way of exposing external knowledge to your agent. You can connect URLs, files, or platform integrations, keep them synced, and toggle availability across environments or variants.
Knowledge connection does not replace the Managed Topics. They work together, and each offers different levels of control and complexity.
Why use multiple knowledge sources?
Your organization may have knowledge living in:
- websites
- documents such as PDFs, CSV, JSON, and other internal reference files
- existing knowledge bases, help desk systems or applications, such as Zendesk or Gladly
Knowledge connection brings these together, allows you to keep them updated, and lets you reuse them across projects without rewriting content.
How Connected Knowledge differs from the Managed Topics
Both the Connected Knowledge and Managed Topics tabs expose information to your agent. They differ in the following ways:
Connected Knowledge
- A connection layer for external knowledge.
- Fast to set up and simple to manage.
- Ideal for FAQ-style agents and large volumes of continuously updated content.
- No prompting skill required.
- Cannot: trigger actions, flows, SMS, hand-offs, or other agentic functions.
- Cannot: specify utterances or control when/why the agent uses specific pieces of information.
Managed Topics
- A curated library of topics and prompts.
- Offers fine-grained control over utterances, behaviours, and what the agent says.
- Can trigger functions, flows, and other agentic actions.
- Requires more time, expertise, and maintenance — but enables anything beyond a simple FAQ bot.
| Feature | Pain points solved | Use cases |
|---|
| Connected Knowledge | Helps teams avoid maintaining another curated knowledge base and gives non-technical users a simple, fast way to connect and manage external data. | Best for FAQ bots, quickly incorporating external knowledge, and controlling data access across environments or variants. |
| Managed Topics | Solves the need for actions, functions, flows, SMS triggers, and offers precise, curated control over agent utterances. | Ideal for agentic behaviour, structured and stable knowledge, and projects requiring complex logic or fine-grained control. |
Add a new source
- Go to Build → Connected Knowledge
- Select New source
- Choose one of:
- Upload files
- Add URL
- Zendesk
- Gladly
- Complete the required details and click Add
Your agent will begin Syncing the content. Once ready, the source appears in the list.
Supported source types
| Type | Examples |
|---|
| Upload files | PDF, TXT, DOCX, CSV, JSON |
| URL scraping | Documentation pages, help centre articles |
| Zendesk (beta) | Help centre + API access |
| Gladly (beta) | Knowledge source sync |
What exactly gets scraped when I upload a URL?
Connected Knowledge uses a third-party scraper that traverses linked pages, but with limits:
- Depth → Only one level below the initial URL.
- Breadth → A maximum of 10 embedded pages.
If your page contains more than 10 links, not all will be scraped. In that case, upload additional URLs individually or use integrations like Zendesk/Gladly for complete coverage.
We recommend to connect applications such as Zendesk, over relying on websites where possible!
Keeping content fresh
After external content changes:
- click Update to re-scrape files or URLs
- or use the Sync icon per source
If a URL requires login or credentials change, syncing may fail. Update access and retry.
Group and manage sources
Group sources by product line, team, region, or document type. Sort by newest, oldest, type, or name. Each source offers:
- Sync
- Rename
- Move to group
- Remove
Why isn’t my agent using the sources I connected?
Several factors affect retrieval:
Data structure
Connected Knowledge splits content into 2000-character chunks with 500-character overlap. Very large documents or widely separated related sections may struggle more with relevance.
What to do:
- Restructure documents into smaller, tighter pieces.
- Repeat key headings or terms.
- Or curate the material as a managed topic for guaranteed usage.
Update state
Two updates must be current:
- Source Update → keeps the data in each source fresh
- Agent Update → applies knowledge connection changes to the agent
Both can be triggered manually. Agent updates also run automatically every few minutes.
Environments, variants, saved changes
Each source must be enabled in the correct environment and variant. Any edits must be saved before leaving the page.
If the Managed Topics and Connected Knowledge contain conflicting data, the Managed Topics wins. Content from Managed Topics is always prioritised
Tips & tricks
- Use PolyAI’s Raven LLM for best results — it paraphrases structured and unstructured content more naturally.
- Connected Knowledge results receive a small bias boost to encourage use; this can be tuned if needed.
- Connected Knowledge data and Managed Topics are merged at runtime.
