Build your agent
Smart Segments
Smart Segments split one agent's knowledge into named
areas β Products, Support, Docs β routed by the
page the visitor is chatting from. A visitor on
/products/β¦ gets product answers first; a visitor in your
help center gets support answers first. The rest of the knowledge base
stays reachable: segments prioritize, they never wall off.
How a conversation gets its segment
- The visitor opens the chat on some page. The page's URL path is matched against every segment's path list β the most specific (longest) match wins; unmatched pages go to the default segment.
- The winning segment is stored on the conversation once, at creation. Navigating to other pages mid-conversation never re-routes it β context stays stable for the whole chat.
- Every answer first searches the conversation's segment. If nothing clears the agent's confidence threshold there, the agent automatically runs one fallback pass across all segments (with a small preference for the home segment on near-ties) before giving up.
- Only when the whole knowledge base has no grounded answer does Strict Knowledge Mode defer β a segment miss alone never triggers the fallback message.
No fallback chains, by design
There is exactly one fallback: segment β everything. Configurable
"try segment B, then C, then D" chains multiply the ways retrieval
can silently do the wrong thing, so they don't exist. Every
cross-segment fallback is recorded on the turn's trace, so you can
see how often visitors needed knowledge from outside their segment.
URL path patterns
- Patterns match the URL path only β query strings and
#fragmentsare ignored; matching is case-insensitive. - A path matches itself and everything under it:
/productsalso covers/products/red-pen, but never/products-old(boundaries are respected). - A trailing
/*is accepted as the same section match:/docs/*β‘/docs. *alone matches every page β useful as an explicit catch-all on a non-default segment.- The longest matching pattern wins across all
segments: with
/products(Products) and/products/manuals(Docs), a visitor on/products/manuals/v2lands in Docs.
Setting it up
- Open your agent β Smart Segments and add segments. The first one you create becomes the default (you can move the star later). Give each segment the URL paths that should route into it; the default usually needs none.
- On the Sources page, assign each knowledge source to a segment with the per-row picker. Assignment is instant β membership lives in the database, so no re-crawl or re-index happens when you move a source.
- Sources you never assign behave as default-segment knowledge, so an agent that predates segments keeps answering exactly as before.
- Rehearse in the Playground:
the Scenario tab gets a Knowledge segment override so you
can ask "what would a
/productsvisitor get?" without opening the live site. The Diagnostics tab shows whether the reply came from the home segment or via the fallback.
What segments do NOT change
- Curated answers, workflows, and the human-handoff / booking shortcuts run before retrieval and ignore segments entirely.
- The widget embed is unchanged β routing happens server-side from the page URL the widget already reports. Nothing to re-install.
- Agents with zero segments behave byte-identically to before the feature existed.
- Deleting a segment re-points its sources at the default segment (deleting the last one turns the feature off); conversations pinned to it simply fall back to whole-knowledge-base retrieval.
Sticky by contract
The segment is chosen when the conversation starts and never
re-resolved β a visitor who wanders from /products to
/support mid-chat keeps the Products context they
started in. If they return after 24 hours of inactivity, a fresh
conversation (and fresh routing) begins.
Plans
The per-plan Smart Segments per agent limit controls the
feature: blank means unlimited, 0 hides it, a number caps
how many segments each agent may define. Platform admins set it under
Admin β Plans.