Last updated: October 22, 2026
We are committed to protecting the confidentiality, integrity, and availability of the information entrusted to PredictHer. This Privacy Policy is intended to explain, in technical and operational terms, what information we collect, how we collect it, how we use it, how long we retain it, how it may be disclosed, and the safeguards we use to protect it.
Sophie is a wellness-oriented product that combines cycle tracking, optional wearable integrations, astrological and lunar features, journaling tools, social features, and AI-assisted personalization. The Services are not intended to provide medical diagnosis, treatment, or emergency services.
This Privacy Policy covers how PredictHer LLC ("PredictHer," "we," "our," or "us"), the operator of the Sophie mobile application ("Sophie" or "the App"), collects and processes personal information obtained through your use of our website, mobile application, and related services (collectively, the "Services") or otherwise provided by you. It should be read together with any additional notice that may be presented at the time specific information is requested.
The Services include the Sophie mobile application, associated website properties, connected integrations, and related support and communication channels. Sophie combines cycle tracking, optional wearable integrations (Oura, Apple Health, Android Health Connect and Samsung Health, Fitbit through Google Health, and WHOOP), activity and journaling features, AI-assisted analysis, lunar and astrological features, and optional social features.
PredictHer is responsible for determining the purposes and means of processing personal information described in this Privacy Policy. Questions regarding privacy practices may be directed to hello@predicther.ai.
We collect the categories of personal information described below to the extent necessary to operate, secure, and improve the Services. Certain categories may constitute sensitive or health-related information under applicable law.
Some of the information described above, including cycle history, fertility-related information, wearable-derived biometric information, and certain journal entries, may constitute sensitive or health-related information under applicable law.
We collect information that you provide when creating an account, completing onboarding, entering period history, adding rituals or journal entries, editing workouts, connecting with friends, contacting support, or otherwise interacting with the Services.
If you elect to connect Oura, WHOOP, Fitbit (Google Health), Apple Health, Android Health Connect, Samsung Health, or another supported data source, we receive the categories of information that you authorize that source to share with us. Depending on the integration and your settings, this may include sleep data, readiness-related data, temperature trends, heart-rate-related data, respiratory data, activity summaries, workout history, and related metadata.
WHOOP. Where you connect WHOOP, we receive only the categories you authorize on WHOOP's consent screen: recovery (including heart rate variability, resting heart rate, blood oxygen and skin temperature), sleep, daily strain and energy, workouts, and basic profile information. We do not request your body measurements. We use WHOOP data solely to provide the features of the Services to you, we do not sell it or use it for advertising, and disconnecting WHOOP in Sophie also revokes Sophie's access in your WHOOP account.
Apple HealthKit. Where you grant Apple Health permissions, we access only the data types you approve, and we use information obtained through HealthKit solely to provide the health and wellness features of the Services. We do not use HealthKit data for advertising, marketing, or similar purposes, we do not sell it, and we do not disclose it to third parties except to service providers processing it on our behalf to provide the Services, or as you otherwise direct.
When you use the Services, we may automatically collect device and usage information reasonably necessary to authenticate sessions, maintain app continuity, debug issues, support push notifications, and understand feature usage. The Services also use local storage, session storage, and similar device-resident mechanisms to preserve app state and preferences.
To understand which features are used and to detect crashes and errors, the Services include a product-analytics component (PostHog) and a crash- and error-reporting component (Sentry). These components receive event names, screen or feature identifiers, basic device and app-version information, and a pseudonymous account identifier. They do not receive your cycle data, symptoms, temperatures or other biometric values, journal entries, or the contents of your conversations with Sophie. Automatic capture of on-screen content, page views, and session recording or replay are disabled. This information is used for product analytics and reliability only, and never for advertising or for tracking you across other companies' apps or websites.
We may receive technical, authentication, geocoding, notification-delivery, infrastructure, or access-status information from service providers that support the Services.
It is inherent in the nature of the Services that we process the information you add to the Services in order to operate them. This includes administering accounts, restoring app state across sign-in and reinstall flows, rendering charts and timelines, enabling journaling and social features, and presenting personalized wellness-oriented content.
We use period history, cycle configuration, prediction history, and, where available, wearable-derived measurements to compute cycle-day context, period status, prediction adjustments, and related wellness-oriented analysis. These outputs may be generated using deterministic rules, statistical logic, or model-assisted analysis.
Where you connect Oura, WHOOP, Fitbit (Google Health), or another wearable source, we use the authorized data to display biometric summaries, activity history, readiness-related information, and related insights within the Services. We also maintain integration state, including source attribution and first-link timing, in order to preserve continuity and analyze feature adoption.
We use personal information to generate AI-assisted chat responses, summaries, forecasts, daily actions, and other in-product analysis. To generate these outputs, relevant information is transmitted to our third-party AI service providers, Anthropic, PBC and OpenAI, L.L.C., acting as processors on our behalf.
What is sent: the messages you write or speak to Sophie (including voice audio and transcripts when you choose to use voice features) and the cycle and wellness context needed to personalize the response — such as your cycle phase and cycle-day context, recent check-ins and logged symptoms, and summaries of wearable-derived measurements (for example sleep, temperature trend, and heart-rate metrics).
Voice. When you talk with Sophie by voice, your conversation is carried in real time between your device, our servers and OpenAI by LiveKit, Inc., our real-time voice provider, acting as a processor on our behalf.
Your permission: the app asks for your consent when you create your account, and again before you chat with Sophie by text or voice if you have not yet answered; those conversations do not begin until you agree. Some AI-assisted features run on our servers rather than inside a conversation — for example, Sophie’s proactive messages, the daily reading, grounding, Oracle and daily actions on your home screen, the overnight organizing of what Sophie remembers from conversations you have already had, and the guidance shown to a partner you have linked. Those server-side features run only after you agree. Until you answer the consent question, or if you choose “Not now,” they pause: none of the context described above is sent to these providers in the background, and the app shows its standard, non-personalized content in their place. Accounts created before the consent question was introduced are asked once more; for a short transition period after this version of the policy took effect they continue as before, and after it they follow the same rule. You can decline or withdraw consent at any time in Settings. Once you decline, our servers stop sending your information to these AI providers, typically within a minute — a voice conversation in progress ends — while the rest of the Services remain available.
How it is protected: these providers process your information solely to generate the requested output under contractual terms that prohibit using your information to train their models or for advertising. Your information is never sold. We may also derive structured personalization features, including self-map or memory attributes, from your interactions with the Services.
When you install the app after seeing one of our advertisements, we use a mobile measurement provider, AppsFlyer Ltd., to understand which advertising campaign led to the installation and whether a trial or subscription followed. The information involved is limited to device and technical identifiers (such as IP address, device model, operating system, an AppsFlyer device identifier, and — on Android, where permitted — the advertising identifier), the fact of installation and app opens, and subscription commerce events (trial started, subscription purchased) forwarded from our subscription-management provider. On iOS, the app also sends AppsFlyer a single “trial started” event at the moment a free trial begins, with no values or other information attached, so that Apple's aggregate campaign reporting can count it. The app sends AppsFlyer no other events.
Subscriptions (RevenueCat). Our subscription-management provider, RevenueCat, Inc., receives your purchase and trial events from Apple or Google to manage your subscription, together with a pseudonymous account identifier and the AppsFlyer device identifier, so that a purchase can be matched to the installation that led to it. On Android, RevenueCat also receives the device's advertising identifier and related device identifiers (such as the Android ID and IP address). RevenueCat forwards these subscription events — trial started, subscription purchased or renewed, with the product and its price — to AppsFlyer, and can forward them to Meta Platforms, Inc. so that Meta can measure our advertising.
Apple Search Ads. On iOS, the app asks Apple's AdServices framework for a short-lived attribution token and sends it to our servers, which ask Apple whether the installation came from an Apple Search Ads advertisement and, if so, which campaign and search keyword. We store that campaign information with your account so we can tell which advertising works. No advertising identifier is involved.
Advertising networks. To measure and improve a campaign, AppsFlyer — and RevenueCat, for Meta — may tell the advertising network that showed the ad (for example Meta or Apple) that an installation, trial start or subscription purchase occurred. On Android, where you have not opted out of ad personalization in your device settings, this may include the device advertising identifier. On iOS, campaign attribution relies on Apple's aggregate SKAdNetwork reports, which do not identify you.
Your health information is never part of this. No cycle data, period dates, symptoms, temperatures, or any other health or wellness information is ever shared with AppsFlyer, RevenueCat, Meta, or any other advertising platform; they receive install and purchase events only. We do not sell your personal information. On iOS, campaign measurement uses Apple's SKAdNetwork framework, which reports in aggregate and does not identify you; we do not request App Tracking Transparency permission and do not use the iOS advertising identifier.
We process technical and operational information to identify bugs, detect abuse, reconcile sync conflicts, investigate failures, maintain availability, and improve the quality and reliability of the Services. We may analyze data in aggregate or de-identified form to better understand product usage and system performance.
We may use your contact and support-related information to respond to inquiries, deliver service messages, resolve technical issues, and notify you of relevant product or policy changes.
We retain personal information for as long as reasonably necessary to provide the Services, maintain account continuity, support restore functionality, comply with legal obligations, resolve disputes, and enforce our terms and policies. Retention periods may vary by data category, by whether the information remains associated with an active account, and by whether the information is stored only on device, synchronized to hosted infrastructure, or both.
Locally cached information may remain on a device until overwritten, removed by the app, or cleared by the user. If you delete your account, we will delete or de-identify cloud-stored account data within a commercially reasonable period, subject to legal, security, anti-fraud, backup, and operational retention requirements.
We may disclose personal information to the following categories of recipients, to the extent necessary for the purposes described in this Privacy Policy:
If you intentionally connect to another user, designate a partner, use a share code, or authorize a connected integration, you direct us to make the data exchanges reasonably necessary to provide that feature. Information shared in those contexts is subject to the feature design and your own choices.
We do not sell personal information for money. We do not disclose personal information to third parties for cross-context behavioral advertising.
This table and the identical list inside the app's Settings are generated from one source file, so they cannot tell different stories.
| Data | Goes to | Never |
|---|---|---|
| Profile and chart | Sophie's servers (your account). With AI consent, your first name and current context reach Anthropic/OpenAI only inside your own conversations. | Sold to anyone. Ad networks. Data brokers. |
| Cycle and reproductive context | Sophie's servers. Your partner sees only the sharing tier you chose. AI providers see it only inside your own chats. If you turn on Amber Alert, the people you add get one text a month: your first name, that your luteal phase has begun, and roughly when your period is likely. | Sold. Advertisers. Employers or insurers. Any third party without an action you took. |
| Wearable signals | Pulled FROM your wearable account into Sophie's servers, for your own analysis. | Sent back to the wearable maker. Sold. Shared with your partner beyond the tier you chose. |
| Sophie memory | Sophie's servers (only while memory consent is on). The AI model serving your chat. | Model training — Anthropic and OpenAI do not train on this API traffic. Anyone else. |
| Partner and friends | Only the specific people you linked, at the tier you set. | Anyone you didn't link. A revoked partner — access ends at revoke. |
| Journals, synchronicities, and rituals | Sophie's servers when synced. Sophie's AI context only if you turned the nightly-factors toggle on. | Your partner or friends. Sale or advertising, ever. |
| Notifications and device state | Apple/Google push services carry each notification to your device. | Marketing platforms. Anyone beyond delivering your own notifications. |
| Purchases and analytics | RevenueCat (subscription state; it passes trial and purchase events to AppsFlyer and can pass them to Meta to measure our ads), AppsFlyer (installs, plus one no-values “trial started” event on iOS; it may pass the fact of an install, trial or purchase to the ad network that showed the ad), Apple (Apple Search Ads attribution), PostHog (product analytics), Sentry (crash logs). | Your cycle, wearable, chat, or journal content — install, purchase and crash events only. |
A Sophie user may add you to her “circle” and send you a personal invitation from her own phone. You receive text messages from Sophie only if you opt in yourself on our consent page at predicther.ai/amber, by entering your own mobile number and confirming. When you opt in, we send one confirmation text; after that you receive at most one text message per month. Message frequency: up to 1 message per month, plus a one-time opt-in confirmation. Message and data rates may apply. Reply STOP at any time to opt out and HELP for help, and the person who added you can remove you instantly in the app.
No mobile information will be shared with third parties or affiliates for marketing or promotional purposes. Mobile numbers and text-messaging opt-in data and consent are used only to deliver the messages described here. They are never sold, rented, or shared with any third party, except our messaging delivery provider (Twilio), which processes them solely to transmit the messages. All other categories of sharing described in this Privacy Policy exclude text-messaging originator opt-in data and consent; this information will not be shared with any third parties.
What the messages say. Amber Alert texts tell the people a Sophie user chooses that her luteal phase has begun, roughly how many days it is likely to last, and roughly when her period is likely (for example, “Her period is likely around Oct 15.”), together with her first name. She chooses who is in her circle; each person consents for themselves, receives at most one text a month, and can reply STOP at any time. The messages never include her symptoms, mood, temperatures, or anything else she logged.
We use administrative, technical, and organizational safeguards designed to protect personal information against unauthorized access, misuse, alteration, disclosure, or loss. Our security controls evolve over time and may vary by system, data class, and threat model.
No security system is perfect. Although we use measures designed to protect personal information, we cannot guarantee that unauthorized access, disclosure, or loss will never occur.
The Services may contain links to, or rely on, third-party websites, APIs, or services. Those third parties operate under their own terms and privacy policies. We are not responsible for the privacy or security practices of third parties except as required by applicable law.
Depending on the feature, you may be able to edit profile information, modify or delete cycle entries, update journals, disconnect any connected wearable, remove social connections, and initiate account deletion directly from within the Services.
Subject to applicable law, you may have rights to access, correct, delete, or obtain information relating to you. You also may have the right to request additional information regarding our processing practices. The rights available to you may vary depending on your jurisdiction.
Because the Services use both local device storage and cloud synchronization, a change may not be reflected in every app surface or device until the next successful synchronization cycle.
If you reside in a U.S. state that provides privacy rights, you may have additional rights, subject to applicable exceptions and verification requirements, including the right to know, access, correct, delete, or obtain a copy of certain personal information, and the right to appeal a denied request where required by law. PredictHer does not sell personal information and does not share personal information for cross-context behavioral advertising.
You may submit privacy-related requests by contacting hello@predicther.ai. We may require information reasonably necessary to verify your identity and protect against fraudulent requests before responding.
We may use or disclose personal information, including sensitive information, for the business and operational purposes described in this Privacy Policy. Those purposes include providing and maintaining the Services, personalization, account administration, security, debugging, analytics, fraud prevention, legal compliance, and related internal operations.
To the extent applicable U.S. state laws treat certain information as consumer health data, PredictHer may collect menstrual and cycle history, fertility-related information, pregnancy-related context, wearable-derived biometric indicators, activity history, and journaled wellness information that you choose to provide. We collect and process that information to provide the Services, generate personalized wellness-oriented features, maintain account continuity, support restore and synchronization flows, and improve and secure the Services.
We collect consumer health data directly from you, from your device, and from connected services that you authorize, such as Oura, WHOOP, Fitbit (Google Health), Apple Health, Android Health Connect or Samsung Health.
We may disclose consumer health data to service providers that operate the Services on our behalf, to connected services at your direction, to users you intentionally connect with inside the Services, and where reasonably necessary for legal, security, fraud-prevention, or transaction-related reasons described in this Privacy Policy.
To the extent provided by applicable law, you may request access to, deletion of, or additional information regarding consumer health data by contacting hello@predicther.ai. We may require verification before responding.
Questions regarding this U.S. Consumer Health Data Notice may be directed to hello@predicther.ai.
The Services are not directed to children under 13, and we do not knowingly collect personal information from children under 13. If we become aware that we have received personal information from a child under 13 without appropriate authorization, we will take commercially reasonable steps to delete that information.
Your information may be processed in the United States and other jurisdictions in which we or our service providers operate. By using the Services, you understand that information may be transferred to and processed in jurisdictions that may have data protection laws different from those in your place of residence.
We may amend this Privacy Policy from time to time to reflect changes in the Services, our data practices, or applicable law. If we do so, we will post the revised version here and update the "Last updated" date above. Your continued use of the Services after the effective date of an updated Privacy Policy constitutes your acknowledgment of the revised policy.
Consumer health data means personal information that identifies a person's past, present, or future physical or mental health status where that term is defined by applicable law.
De-identified data means information that has been processed so that it is not reasonably linkable to a specific individual, subject to applicable law and technical limitations.
Personal information means information that identifies, relates to, describes, or can reasonably be linked to an individual.
Self map means structured personalization information derived from your interactions with the Services and stored in your profile to improve continuity and relevance of the user experience.
If you have questions regarding this Privacy Policy or our privacy practices, please contact us at hello@predicther.ai.