Knowledge
Sync Notion, Zendesk, Confluence, Drive, commerce systems, and other sources so assistants read current content—not one-off exports.
Connectors run scheduled or event-driven sync jobs authorized via OAuth or API tokens you provide. Retrieved documents land in knowledge bases the same way uploads do, keeping RAG pipelines unified.
Per-integration setup, scopes, and caveats are documented under Integrations → [product].
Least-privilege OAuth scopes reduce risk during reviews. Use dedicated integration accounts when possible so offboarding an employee does not break sync.
Rotate tokens on the same schedule as other third-party credentials.
A connector can feed one or more knowledge bases; agents then attach those bases. Avoid wiring all-company wiki spaces into customer agents.
When content is multilingual, confirm the right locales sync and that instructions tell the model how to respond per language.
Dashboards should surface sync errors (auth expired, rate limits). Alert owners before customers notice stale answers.
Temporary vendor outages may backlog jobs; replays should be idempotent.
Continue exploring
Platform features · HTTP API & webhooks · Quickstart · Contact
Build on your stack
Start a trial or talk to us about rollout, governance, and enterprise requirements.