How Settld handles your data
This is a factual description, not a reviewed legal policy. It accurately describes what the software does today. Before accepting real customers you need a privacy policy drafted or reviewed by a lawyer for your jurisdiction, covering the legal bases, retention periods, sub-processors, and data-subject rights that apply to you.
What is stored
- Transcripts you upload. The original file and a normalised plain-text version.
- Extracted decisions. The decision text, owner, deadline, confidence score, and a reference to the exact span of transcript it came from.
- Your account. An email address and the workspace it belongs to. No password is stored, because there isn't one.
Where it is stored
Transcript files are held in private object storage and are never served from a public URL — they are fetched server-side after an ownership check. Decisions and account records are held in a managed Postgres database. Both are encrypted at rest by the hosting provider.
Who can see it
Isolation between workspaces is enforced by the database, not by application code remembering to filter. Every query runs scoped to one workspace, and a query that fails to declare its scope returns nothing rather than another workspace's data.
Processing by third parties
- Anthropic — transcript text is sent to the Claude API to extract decisions.
- Voyage AI — decision sentences (not transcripts) are sent to generate the embeddings used for conflict detection.
Your transcripts are not used to train models.
Deletion
Deleting a meeting removes its transcript, its normalised text, its extracted decisions, and the stored file. Deleting a workspace removes everything belonging to it.
Contact
Questions about this: hello@example.com.