Guide 10 of 10
Data and Security Overview
How Mnemo handles your data, encryption, and access controls.
Your data stays in your environment
The most important thing to understand about Mnemo's security model is that your documents and queries are never sent to external AI providers. All processing — indexing, search, response generation — happens within your deployment environment. This is an architectural guarantee, not a policy statement.
What data Mnemo stores
Document content
The text extracted from your uploaded files. Stored in your deployment environment, encrypted at rest.
Vector embeddings
Mathematical representations of your document content used for search. Stored separately from the original text, in your environment.
User accounts
Name, email, organisation, and role. Used for authentication and access control.
Query and conversation history
Questions you have asked and the answers returned. Used for conversation context and audit purposes.
Usage metadata
Timestamps, document counts, query volumes. No document content.
Encryption
In transit: All data is encrypted using TLS 1.2 or higher as it moves between your browser and Mnemo's servers.
At rest: All stored data — documents, embeddings, user data, logs — is encrypted using AES-256.
No training on your data
Your documents and queries are never used to train or fine-tune any AI model. This applies to Arion Flow and to any AI model providers used in the processing chain. It is a contractual and architectural commitment.
Access logging
All logins, document uploads, queries, and permission changes are logged with a timestamp and user identifier. Logs are available for export by organisation administrators and can be used for compliance, audit, and FOI purposes.
More detail
For more detailed technical and compliance information, see these resources: