Skip to main content
Currently available connectors:
  • SharePoint
  • Jira
  • Confluence
We distinguish between semantic connectors and API interfaces to external systems where data is stored. In terms of output quality, a semantic connector is often preferable, but this depends on the individual case.
FeatureSemantic Connector (for RAG)API Connector
Core PrincipleFinds contextually relevant answers from a knowledge base.Queries data live from an external system.
Data ProcessingRequires Indexing: Content is analyzed & stored in a vector database.No Indexing: Works directly with the live data source.
Type of SearchSemantic Search: Understands the meaning and context of a query.Direct Query: Uses the search function provided by the system’s API.
Data SynchronizationSyncs are necessary to keep the knowledge base up-to-date.No Sync required as the connection is real-time.
Data FreshnessData is as current as the last sync.Always real-time data.
AvailabilityHighly Available: Queries can still be served from your local index even if the original source system is temporarily offline.If the external API or network is down, the connector will fail. It is highly dependent on the source system’s availability.
Ideal for…Large volumes of unstructured data (e.g., PDFs, Confluence, Word documents).Structured data or when real-time information is critical (e.g., inventory levels, ticket statuses).
Do you have any relevant knowledge stored in systems that we don’t yet connect to? Talk to us! We would happily discuss the implementation with you. With the help of our standard connector architecture and the use of tool calls as part of our agentic-first roadmap, adding your favorite data is as easy as never before.