Documents
Local workspace notes are saved as Markdown files in your chosen folder.
Understand how documents and knowledge graphs are stored, what reaches an AI model, and how to export or delete data.
Local workspace notes are saved as Markdown files in your chosen folder.
Entities and relationships are stored as a graph linked to source documents.
Domain vocabulary, entity types, properties, and review decisions are stored separately.
Maps and summaries of related topics are saved in the workspace.
Keyword and vector search indexes are stored in the environment you use.
Conversations and the tool activity within them are stored with a link to the workspace.
Your request and the document or conversation context needed for the task are sent to the selected model.
Document and query text is sent to an embedding model to help find similar content.
Recorded audio is sent to the selected transcription model when you use voice features.
Select source documents, the graph, vocabulary, decisions, and other items to export as a ZIP bundle.
You can open local workspace Markdown notes in other editors and copy the files.
Choose which graph data, indexes, and records to clear per workspace, separately from deleting source files.
Cloud workspace documents, graphs, and conversations are stored on Consilience hosting, separately from desktop local storage.
Hosting with backups configured periodically backs up changed workspaces, with included data and retention governed by the operating policy. Agent learned memory and code indexes are excluded from the default backup.
An owner can delete a workspace after confirmation, while server copies and backup retention are managed separately.
Consilience does not use customer data to train models, while data sent to external models is subject to each provider’s retention and training policies.