Getting started
How Claura is structured
Claura separates connected sources, document processing, financial intelligence, and workspace access so every answer remains explainable.
1. Collection
OAuth-connected sources run historical and incremental syncs. Claura records what a sync saw, imported, failed to import, and which cursor should be used next.
2. Processing
Documents are classified and extracted into normalized financial fields. Original artifacts remain available, while processing results and review states are stored separately.
3. Financial reasoning
Search plans a response across the active workspace, reports coverage, and returns answer blocks such as summaries, comparisons, charts, evidence, downloads, and follow-up questions.
4. Workspace isolation
Every workspace-scoped request is authorized against membership and role. Client relationships do not silently grant ownership, and the backend enforces capabilities even when privileged service access is used.
Financial context should stay connected.
Claura keeps the document, conversation, and decision together—so the next person never has to reconstruct the story.
found in Claura documentation · Financial memory