Skip to main content
Unstructured Foundation is an Early Access product. Unstructured must accept you into the Early Access program before you can begin using Foundation. Add your name to the Early Access interest list.Unstructured is currently accepting a limited number of participants into the Early Access program at its discretion, with plans to expand access to a more general audience in the future.The following information is being provided to give you an advance preview of possible future functionality. Unstructured makes no express claims as to what extent this information will reflect the state of the product upon initial general availability.

​
What Foundation storage is

Foundation storage is a per-user database. It holds your enrichment index (the structured, searchable representation of your files). This makes the enrichment index persistent across every session and every AI tool you use. Foundation never moves or replicates your source files. Foundation stores only the derived representations it needs to answer queries. Foundation processes each file once at ingest time. Every subsequent query returns the same quality without reprocessing.

​
What is stored

Foundation stores the following for each file:
  • Named entities (NER): People, organizations, locations, and other entities extracted from file content.
  • Topics: Subject labels generated from file content.
  • Short summaries: File-level summaries generated at ingest time.
  • Full text: The complete text of the file, stored in structured form.
  • File metadata: Title, source, MIME type, creation date, and last-modified date.
Your source files remain at rest in their original connected locations. Foundation does not copy, move, or access them after ingestion.

​
Where it is stored

Foundation stores your enrichment index in a dedicated database provisioned exclusively for your account. Your database is isolated from all other users; Foundation shares no data across accounts. Every query your AI tool makes returns results from your index only. You do not access this database directly. All queries go through the Foundation MCP server, which verifies your identity before executing any operation. Within your index, Foundation organizes data by source. A source is a configured connection to a data source. Each one contributes its own files to your index. When you query Foundation, your AI tool can search across all sources at once or narrow to a specific one.

​
Viewing your stored data

There is no raw data browser. To see what Foundation has indexed, ask your AI tool: What does Unstructured Foundation have access to? Your AI tool returns a summary of indexed sources, file counts, and when each was last updated. You can also ask about a specific source, for example: How many files does Unstructured Foundation have from Google Drive?

​
Deleting stored data

When you remove a source, Foundation prompts you to delete the indexed data associated with it. If you confirm, Foundation removes all stored representations for files from that source. Your source files in their original connected location are not affected.

​
Next steps