Privacy Policy
Last updated: September 25, 2026
o1o1 International, LLC ("Company", "we", "us") operates Persona. This Privacy Policy explains what data we collect, how we use it, and your rights. Persona is designed with a privacy-first architecture: your raw content stays on your device.
1. Data We Collect
Data that stays on your device (never uploaded)
- File contents indexed by the local Persona Node application
- Email bodies read from your local Outlook database
- Browser history (if enabled)
This data is stored in a local SQLite database on your computer. We have no access to it. It is never transmitted to our servers.
Data stored on our cloud
| Data | Purpose |
|---|---|
| Account info (email, name) | Authentication, billing |
| Knowledge graph facts (structured triples) | Cross-agent memory, semantic search |
| Entity names and aliases | Entity resolution across sources |
| Embedding vectors | Semantic search (GraphRAG) |
| Conversation logs (user input + agent response) | Cross-session memory continuity |
| Metadata summaries (titles, topics, people) | Search and knowledge extraction |
| Behavioral/communication-style summaries (LLM-generated narrative of your working style, preferences, and recent activity) | Personalizing responses across sessions |
| Billing and subscription data | Payment processing |
Data sent to third-party LLM providers
To extract structured knowledge from your content, we send text to LLM providers for processing. This includes:
- File content snippets (for fact extraction)
- Email subjects and summaries (for enrichment)
- Conversation text (for fact extraction)
We use OpenAI and/or Google Gemini for this processing. Their privacy policies apply to data sent to them. We do not send your raw files or full email bodies — only the content necessary for enrichment (truncated to ~8,000 characters per item).
We also send conversation text to TypeSafe AI, whose model decides whether a conversation contains information worth remembering. TypeSafe returns only a probability and does not generate content. Text sent to TypeSafe is truncated to ~4,000 characters with secrets redacted.
Sensitive PII Handling
Before content is stored in our cloud database, we automatically detect and tokenize high-risk personal identifiers. This is a pseudonymization measure (GDPR Article 4(5)) applied in addition to transport and database encryption.
Automatically tokenized (raw value discarded, not stored):
- US Social Security Numbers (SSN)
- Credit card numbers (Luhn-validated)
- Bank account numbers (when clearly labeled)
- IBAN (international bank account numbers)
- Passport numbers (when clearly labeled)
When detected, the raw value is replaced with an opaque token and immediately discarded. We retain no copy of the original value — there is no vault, decryption key, or reveal mechanism. Retrieval of the raw value through our service is not supported.
Audited but not tokenized:
- Phone numbers
- Email addresses
These are logged for compliance auditing but remain inline in your stored content.
Not currently detected:
Our current detection is pattern-based. Categories that require contextual understanding — names, physical addresses, medical records, and GDPR special categories — are not currently tokenized. Users handling such categories should exercise additional discretion.
What this means for you:
- LLM processing sees raw content: Tokenization happens before cloud storage but after AI enrichment. LLM providers invoked by your agent may temporarily process raw PII, subject to their own privacy policies.
- Cloud database stores tokens only: High-risk PII does not persist in raw form on our servers.
- Local device keeps raw values: Chat transcripts on your Persona Node device retain original text (not uploaded).
- Retrieval is not supported: If you upload a document containing your SSN and later need to reference it, you must retrieve it from your original source — raw values cannot be returned by our service.
2. How We Use Your Data
- Provide the service: index content, extract facts, enable semantic search, maintain cross-agent memory.
- Billing: process subscriptions and payments via Stripe.
- Improve the service: aggregate, anonymized usage metrics (no personal content).
We do not:
- Sell your data to third parties.
- Use your content to train AI models.
- Share your data with advertisers.
- Access your raw files or email bodies (they stay on your device).
- Store raw values of detected high-risk personal identifiers (SSN, credit cards, bank accounts, IBAN, passports) on our servers.
3. Data Storage & Security
- Cloud data is stored on Google Cloud Platform (Cloud SQL, GCE) in the US.
- All connections are encrypted via TLS.
- Database connections use SSL.
- Authentication via Firebase Auth (Google).
- Billing handled by Stripe (PCI-DSS compliant).
- High-risk PII is pseudonymized (GDPR Article 4(5)) before database storage — see Sensitive PII Handling above.
4. Your Rights
- Access: You can view all your stored facts and conversation history through the memory tool. High-risk PII identifiers appear as tokens in exported data; retrieval of the raw values is not supported through our service (see Sensitive PII Handling).
- Export: You can export your knowledge graph data at any time.
- Delete: You can delete individual facts, conversations, or your entire account. Cloud data is permanently removed within 30 days of account deletion.
- Portability: Your data is accessible via the standard MCP protocol and can be used with any compatible AI agent.
- Local data: Local data on your device is always under your control. We cannot access, modify, or delete it.
5. Data Retention
- Active account: Data is retained as long as your account is active.
- Canceled subscription: Cloud data is retained for 30 days after cancellation, then permanently deleted.
- Account deletion: All cloud data is permanently deleted within 30 days.
- Local data: Local data on your device persists until you manually delete it or uninstall the Persona Node application.
6. Third-Party Services
| Service | Purpose | Data shared |
|---|---|---|
| Google Firebase | Authentication | Email, password hash |
| Stripe | Billing | Payment info, email |
| OpenAI | LLM enrichment | Content snippets for fact extraction |
| Google Gemini | LLM enrichment (fallback) | Content snippets for fact extraction |
| TypeSafe AI | Memory filtering | Conversation text (up to ~4,000 characters, secrets redacted) |
| Google Cloud Platform | Infrastructure | Cloud-stored data (encrypted) |
7. Children
Persona is not intended for use by anyone under 18. We do not knowingly collect data from children.
8. Changes to This Policy
We may update this Privacy Policy from time to time. We will notify you of material changes via email or in-app notification. Continued use after changes constitutes acceptance.
9. Contact
Privacy questions, data deletion requests, or requests to purge content that may contain sensitive PII: [email protected]
o1o1 International, LLC
8118 NE 142nd Pl, Kirkland, WA 98034