Privacy Policy

We collect what you give us — and only what you give us. There are no third-party trackers running on our app surfaces, and we do not buy data about you from data brokers.

• Account information: name, email, password (hashed), authentication method (email, Google, Microsoft, Apple, or passkey), and the role you signed up under.
• Profile content: profile name, age (if you provide it), pronouns, diagnoses, support needs, accommodation notes — whatever you choose to enter into the profiles you create or manage.
• Plan content: plans, goals, milestones, tasks, tracker entries, calendar items, timeline events, activity feed posts, message threads, and files you upload.
• Navi conversations: Navi is available to verified professionals and platform admins. If you use it, we save your messages to Navi and its responses to your account so you can return to them. Navi also saves short notes from these conversations, such as a preference or a decision, to the profile you were working on, and uses them in later Navi conversations about that profile, within the limits section 3 describes. To find the right notes, the text of each one is turned into a numeric representation by Google's Vertex AI text-embedding model. Messages and notes are encrypted at rest. Plans made with Navi's help are ordinary plans and stay yours. On accounts that aren't verified professional or platform admin accounts, Navi conversations from before September 27, 2026 aren't shown in the app, and we permanently delete them on December 26, 2026. Backup copies expire on our normal backup schedule.
• Sharing graph: who you've invited, what they can see, and the audit trail of grants and revocations.
• Operational data: request logs, error reports, performance metrics. These logs do not include the content of plans, goals, notes, messages, or other fields you fill in. A CI drift-guard test blocks the common log patterns that would interpolate user content; a handful of multi-line cases are audited manually until the guard is upgraded.

Any of the content above may include health-related, behavioral, educational, or developmental information if that's what you're organizing. We treat that content as sensitive in the sections that follow.

Your information is used to provide and improve the service you signed up for: organize your plans, surface the right resources, keep your team coordinated, send service-critical notifications, and run the platform safely.

We do not sell your information. We do not rent your information. We do not use it to train AI models — neither ours nor any vendor's. We do not run advertising and we do not share data with advertising networks.

Navi is Specthrive's AI assistant for verified professionals and platform admins. It answers questions about the plans they work on, drafts and edits plans and guides, breaks tasks into smaller steps, suggests goals from resources, helps draft and summarize care-team messages, and adds suggestions and explanations to their Smart Search results. Navi runs on Google Cloud's Vertex AI, using Google's Gemini models. Navi's requests use Vertex AI's global endpoint, so Google may process them outside the United States.

When a professional uses Navi on a profile that has been shared with them, Navi reads only the areas that share covers. If the share covers plans, Navi reads the person's first name, age, stage, state, interests and strengths; their active diagnoses; their plans, goals, milestones and tasks, with titles, descriptions, areas and progress, including which tasks are overdue; and notes Navi saved about that profile in that professional's own earlier chats. If it covers trackers, Navi reads which trackers were used recently, how often and when, but not what was entered. If it covers documents, Navi sees a document's title, category, file type, detected document type and text-recognition status, never its contents. In any conversation, Navi also rereads that professional's earlier messages and its own replies. A professional who manages the profile, or belongs to the workspace that manages it, has Navi read all of these, including notes Navi saved to that profile earlier.

If the professional also manages other profiles, Navi sees their first names and how many open tasks each has. When a professional who has access to reflections asks for a plan draft from a Reflection Map, the reflection text is sent too. When a professional drafts or summarizes a care-team message with Navi, the last ten messages of that conversation are sent. When a professional runs Smart Search with a profile selected, their search text and that profile's details, including its diagnoses, are sent too. For a profile shared with them, the profile's details are sent only when the share covers plans; without it, the search uses their text alone.

While answering, Navi can look things up on Google Search. When it does, it sends Google a search query it writes, which may reflect the conversation. Vertex AI does not use your data to train Google's or anyone else's models. Any caching or short-term retention within Vertex AI is governed by Google Cloud's published data-handling policies. Your data is not sold. Apart from those search queries, it is not shared with parties beyond Google Cloud's processing infrastructure.

AI that works for everyone. Two features use AI for every adult account. Search: when you search guides, or use Smart Search on Discover, your search text is turned into a numeric representation by Google's Vertex AI text-embedding model, in the United States, so we can find close matches. When you search with a profile selected, the profile's age group and up to two of its diagnoses are included. For a profile shared with you, they are included only when the share covers plans. Document text: when you upload a PDF or image, Google Cloud Document AI reads the text in it, in the United States, so your documents are searchable. Apart from Navi, suggestions, recommendations and reminders in Specthrive come from rules we write, not from an AI model.

Your switches. The person who manages a profile can switch AI features off for that profile from Settings → Privacy → AI features. The switches cover Navi conversations about the profile; Navi's goal suggestions, task breakdowns and plan drafts from Reflection Maps; and profile-aware search, search suggestions and recommendations. Drafting and summarizing care-team messages with Navi, search without a profile selected, and document text recognition don't have their own switch yet.

Navi suggests; the professional chooses what to act on. Navi never sends a message or changes a plan without their explicit confirmation. AI features, including Navi and search, are not available to account holders under 18, and documents on an under-18 account holder's own profile are not sent for text recognition.

Specthrive uses a layered consent model:

All consent decisions (grants, revocations, acknowledgments) are audit-logged for accountability. Audit records do not include the underlying personal information.

You decide who sees what. Sharing is per-profile and per-domain — you can give a coach access to plans and trackers, give a babysitter access only to the calendar, and keep documents and messages off-limits to everyone except yourself. Documents are never shared by default.

When you share, we record what was granted, who granted it, and when. You can revoke access at any time; the revocation is also logged. People you share with see only what their grant allows; they don't see audit logs or the rest of your team. When a professional uses Navi on a shared profile, Navi reads only the areas the grant covers, as section 3 describes. See Privacy & sharing: who sees what for the full breakdown.

Specthrive runs on a small set of trusted infrastructure providers. We don't share your information beyond what these providers need to deliver the service.

Data in transit is protected with TLS. Within our infrastructure, messages and uploaded documents are encrypted at rest using Fernet symmetric encryption with per-environment keys held in Google Cloud Secret Manager. Database backups inherit the same key boundary.

Sensitive operations are audit-logged with the actor, action, target, IP address, and timestamp. Operational logs do not include user-entered content or authentication tokens. A CI drift-guard test fails on the common log patterns that would interpolate a request body, query params, or an auth header; remaining edge cases (multi-line log calls) are audited manually until the guard handles them. Multi-factor authentication is available on every account.

Specthrive is not a HIPAA-covered entity today and does not claim HIPAA compliance. We are not a clinical service, we do not bill insurance, and we do not act as a business associate to a covered entity through the consumer product. We process information that people choose to enter — including health-related notes — and we treat it carefully: encryption, access scoping, audit logging, no PHI in logs, no AI model training on user data, and a BAA in place for our cloud infrastructure.

We are building toward clinical-grade compliance as the product grows into clinic workspaces, and we will surface that posture transparently as it ships rather than asserting compliance we haven't earned.

Specthrive is intended for use by adults coordinating care — including for children, dependents, and others they support. Children should not create their own accounts. When you create or manage a profile for a child, you are the controlling party for that profile's data and we record your guardian attestation accordingly.

We do not knowingly collect information directly from children under 13. If you believe we have, contact privacy@specthrive.com and we'll delete it promptly.

You can:

Residents of California and other US states with comprehensive privacy laws have additional rights to know, access, correct, delete, opt out of profiling, and opt out of the sale or sharing of personal information. We do not sell or share your information for cross-context behavioral advertising. To exercise the other rights, the controls listed in Section 10 are the fastest path; you can also email privacy@specthrive.com.

When we publish a materially updated version of this policy, the app will surface a re-acknowledgment prompt the next time you sign in. Non-material updates (typo fixes, clarifications without behavior change) update the version number but don't prompt for re-acknowledgment. The current version and effective date are at the top of this page.

Specthrive is operated by Specthrive, Inc. — the company responsible for the information described in this policy (the data controller), and the same entity named in our Terms of Service. You can reach us here: