Data Handling for Teams

Elephas in team and enterprise environments

Elephas is designed for professionals working with confidential documents. Whether you are a solo consultant or part of a larger team, the privacy model stays the same: your data is indexed and stored locally on each device.

Data boundaries

• 100% local indexing (default since v11.6): When users add documents to a Super Brain, all indexing happens on their Mac. Document content is not sent to any external service to build the index.
• Smart Redaction (new in v11.7): Personal information is anonymized on-device before any query reaches a cloud AI provider.
• Vector and chunk storage is always local: The search index and text chunks for each Super Brain live on the user's Mac, never on Elephas servers.
• iCloud sync is opt-out: Elephas syncs Super Brains across a user's Apple devices via iCloud Drive by default. Teams that need strict device isolation can disable iCloud sync in Preferences → General.
• AI queries are scoped: When users ask a question, only the relevant text passages (not the full document library) are sent to the AI provider, after Smart Redaction has anonymized any personal information.

For teams evaluating Elephas

Common questions from IT and security teams:

• Where is data stored? The local search index and text chunks live on each user's Mac in ~/Library/Application Support/Elephas/. No centralized server holds your data.
• Does Elephas have access to our data? No. Elephas (the company) never stores your document content, search index, or chat history on its own servers.
• Does any text leave the device during indexing? No, not in version 11.6 and later. Indexing is 100% local on the Mac. No document text is sent to any external service to build the index.
• What about AI queries? When using cloud AI, the user's question and relevant passages are sent to the AI provider (OpenAI, Anthropic, Google). With Smart Redaction (v11.7+) enabled, personal information is anonymized on-device before sending. API data is not used for training. Offline mode keeps even queries local.
• Can we use our own API keys? Yes. Bring Your Own Key mode sends requests directly from the user's device to the provider, with no intermediary.

Recommended configuration for sensitive work

1. Use Offline AI for maximum privacy (no data leaves the device)
2. If cloud AI is needed, enable Sensitive Data Protection to redact PII
3. Use Bring Your Own Key to eliminate the reverse proxy from the data path
4. Keep workspaces focused: one workspace per project to limit the scope of data sent in queries