Buff Kitty
Privacy Policy
Effective date: July 20, 2026. Last reviewed: July 20, 2026.
Summary
Buff Kitty is a protein-tracking app. We try to keep the product simple: core logs and profile data live on your device; server processing is mainly for paid AI food analysis, subscriptions, analytics, security, and support.
Protein, weight, food, and fitness information can be personal or sensitive depending on where you live. Treat this app as a convenience tool, not as a medical record system.
This policy covers the Buff Kitty iOS app and these legal webpages. It does not cover third-party websites or services that we do not control.
Who is responsible
The operator of Buff Kitty is Arta Ghasemi, located in Maryland, United States.
Privacy contact: artaghasemi06@gmail.com. Postal contact, if required by law: No postal address is published; please contact us by email..
Information we collect or process
The table below reflects the audited production app behavior as of the last review date above. PostHog session replay is disabled in production builds in the reviewed app configuration. If the app adds accounts, cloud sync, email marketing, health integrations, session replay, more attribution providers, or more analytics, this policy must be updated before release.
AI food analysis
If you use AI logging, the app sends your typed food description, voice transcript, or selected meal photo to our backend so an AI model can estimate protein. Photos are downsampled and compressed before upload. The backend also receives a per-install device ID and a RevenueCat app user ID to verify paid access and apply rate limits.
If you use voice logging, the app uses iOS Speech Recognition to create a transcript. Buff Kitty's backend receives the transcript text when you choose to analyze it, not a stored raw audio file from Buff Kitty.
Buff Kitty may use OpenAI, Google Gemini, or another comparable model provider for AI food analysis, and we may switch models or providers over time to improve quality, cost, reliability, or safety. These providers receive the food description, voice transcript, or meal image only when needed to return a protein estimate.
Buff Kitty's current backend is designed not to store raw food photos, raw food descriptions, or voice transcripts in its database. It may keep operational logs, error categories, rate-limit counters, status codes, and usage metadata. AI model providers may process and retain inputs and outputs according to their own terms and the service configuration we use.
Do not submit meal photos that include faces, documents, addresses, children, medical records, or anything you would not want processed by an AI provider. The AI estimate can be wrong and should be reviewed before you log it.
Analytics and identifiers
Buff Kitty uses PostHog for product analytics. The current app identifies analytics events using the RevenueCat app user ID, not a name or email address. Event properties can include protein grams, selected onboarding answers, referral or promo code if you provide one, AI result summaries, paywall events, and goal/streak summaries.
The reviewed production app configuration disables PostHog Session Replay. Internal debug builds may use replay for development troubleshooting, but the production app does not offer a user-facing replay opt-in and does not intentionally record production user sessions.
Buff Kitty also uses Apple's AdServices framework and RevenueCat to measure Apple Ads campaign attribution. This uses a short-lived Apple-provided attribution token and does not collect IDFA. Apple controls how much campaign detail is returned based on system privacy settings.
The current app does not include an advertising SDK, does not use IDFA in the audited code, and is not designed to sell personal information or share it for cross-context behavioral advertising. If this changes, the privacy policy, App Store privacy answers, and any required consent flows must change before release.
Device permissions
Camera and photo library
Camera access is used when you choose to photograph a meal for AI protein estimation or scan a packaged-food barcode. Photo library access is used only when you choose a meal photo. Buff Kitty does not need access to your full photo library to function.
Microphone and speech recognition
Microphone and speech recognition access are used only when you choose voice logging. Voice input is transcribed so the transcript can be analyzed like a typed food description.
Notifications
Notification permission is used for local daily reminders. The audited app schedules reminders on the device and does not send a push notification token to Buff Kitty servers. The app may track whether notification permission was granted, denied, or skipped.
How we use information
- Provide protein tracking, barcode lookup, streaks, reminders, paywalls, subscriptions, and AI logging.
- Calculate and display your daily protein target and progress.
- Verify subscription access and prevent abuse of paid AI features.
- Measure product funnels, Apple Ads attribution, reliability, and feature usage.
- Respond to support, privacy, legal, and billing requests.
- Comply with law, enforce terms, and protect Buff Kitty, users, and service providers.
Retention
- On-device app data stays on your device until you delete it, uninstall the app, or use any production deletion control we provide.
- Raw AI photos and descriptions are not intentionally stored by Buff Kitty's database, but may exist transiently in request processing, infrastructure logs, or AI provider systems according to provider settings.
- PostHog analytics events are retained according to the active PostHog project settings, plan, and deletion workflows.
- Subscription, analytics, paywall, support, security, and rate-limit data may be retained as long as needed for the purposes described in this policy, legal obligations, security, dispute resolution, and business records.
- We periodically review retention practices and delete or de-identify information when it is no longer needed for the purposes described in this policy, unless we need to keep it for legal, security, dispute-resolution, or business-record reasons.
Your choices
- You can avoid AI processing by using manual logging instead of text/photo AI logging.
- You can control camera, photo library, and notification permissions in iOS Settings.
- You can cancel or manage subscriptions through your Apple account subscription settings.
- You can remove local app data by deleting the app from your device, unless a future in-app deletion control is provided.
- You can contact us to request deletion of PostHog analytics records that can reasonably be located using your RevenueCat app user ID or other information you provide.
- You can contact artaghasemi06@gmail.com to request access, correction, deletion, or other privacy rights that apply in your region. We may need information such as your RevenueCat app user ID to find vendor-side records.
Regional privacy rights
Depending on where you live, laws such as the CCPA/CPRA, GDPR, UK GDPR, and other state or national privacy laws may give you rights to know, access, correct, delete, restrict, object, port data, withdraw consent, or appeal a privacy decision.
Buff Kitty does not currently sell personal information or share it for cross-context behavioral advertising as those terms are commonly used in US state privacy laws. If you believe otherwise, contact us.
Children
Buff Kitty is not directed to children. You must be at least 18 years old to use Buff Kitty.
Security
We use reasonable technical and organizational measures for a bootstrapped app, including server-side AI keys, subscription checks before paid AI requests, private rate-limit tables, TLS-protected network requests, and platform permission controls. No online service can guarantee absolute security.
Changes
We may update this policy as Buff Kitty changes. If the changes are material, we will update the effective date and provide notice as appropriate in the app, on this site, or through another reasonable method.
Contact
Questions or requests: artaghasemi06@gmail.com.
Current data inventory
This inventory is included directly in the policy so App Store review, users, and legal reviewers can compare the copy against the actual product.
| Category | Examples | Where it goes | Why |
|---|---|---|---|
| Protein and fitness profile | Body weight, training goal, activity level, unit system, daily protein target, reminder time | Stored on your device in SwiftData/UserDefaults. Some segmentation fields are also sent to Superwall and analytics when you finish onboarding. | Calculate your daily protein target, show app state, personalize paywalls, and understand product funnels. |
| Food logging data | Protein grams, timestamps, log source, food names, estimated flag, AI confidence, components count | Stored on your device. Event summaries are sent to PostHog. Raw diary history is not uploaded to Buff Kitty servers by default. | Run the core tracking experience, streaks, reminders, and product analytics. |
| Barcode product lookup | Scanned product barcode, product name and brand, serving size, protein values | The scanned barcode is sent directly to Open Food Facts with routine network metadata and the Buff Kitty app version. Open Food Facts returns public packaged-food product and nutrition data. Buff Kitty stores only the food name and protein amount if you choose to save the result as a local food log. | Look up packaged foods and calculate an editable protein amount without requiring AI analysis. |
| AI food analysis inputs | Short food descriptions, voice transcripts, speech audio used for dictation, compressed meal photos, image MIME type, RevenueCat app user ID, per-install device ID | Speech audio is processed by iOS Speech Recognition when you use voice logging. Transcripts, typed descriptions, and compressed photos are sent to a Supabase Edge Function, then forwarded to OpenAI, Google Gemini, or another comparable AI model provider for estimation. Buff Kitty does not intentionally store raw photos, descriptions, or transcripts in its database. | Transcribe voice input, estimate protein for paid AI logging, verify entitlement, prevent abuse, and control infrastructure cost. |
| Subscription and purchase data | Apple transaction status, RevenueCat app user ID, entitlements, offering and product metadata, subscription status | Processed by Apple, RevenueCat, and Superwall. Buff Kitty uses RevenueCat and Superwall SDKs to show paywalls, process purchases, restore access, and gate paid AI features. | Provide subscriptions, enforce paid access, restore purchases, and support subscription status in the app. |
| Analytics and product events | App lifecycle events, onboarding steps, referral or promo code if you provide one, notification permission result, paywall events, AI analysis start/completion/failure, food logged summaries, goal changes, goal hit summaries | Sent to PostHog and identified with the RevenueCat app user ID so subscription and analytics events can be correlated. | Measure onboarding, paywall conversion, core feature usage, reliability, and product improvements. |
| Apple Ads attribution | Short-lived Apple Ads attribution token, campaign and ad-group attribution metadata | The app asks Apple's AdServices framework for an attribution token and allows RevenueCat to use it to retrieve Apple Ads campaign attribution. The returned detail depends on your Apple privacy settings. Buff Kitty does not use this flow to collect IDFA. | Measure which Apple Ads campaigns lead to installs and subscription revenue. |
| Device, app, and network metadata | App version, OS version, device model, locale, time zone, IP-derived region, request metadata | May be processed by PostHog, Superwall, Supabase, Vercel, RevenueCat, Apple, Open Food Facts, and AI providers as part of SDK, hosting, product lookup, security, and abuse-prevention operations. | Operate the service, debug issues, prevent abuse, localize purchases, and maintain security. |
| Support communications | Email address, message contents, attachments you choose to send | Collected when you submit in-app feedback or contact Buff Kitty directly. In-app feedback is stored in Supabase and may be included in an optional founder webhook if configured. | Respond to support, privacy, legal, and billing questions. |
Service providers
Review each provider before launch and again whenever an SDK, backend, or AI provider changes.
| Provider | Role | Relevant data | Policy |
|---|---|---|---|
| Apple App Store, StoreKit, AdServices, and Speech Recognition | App distribution, in-app purchases, auto-renewable subscriptions, Apple Ads attribution, review prompts, and voice dictation | Apple account, purchase, billing, device information, Apple Ads attribution data, and speech audio/transcription data handled by Apple under Apple's terms when you use platform services. | Review |
| RevenueCat | Subscription infrastructure and entitlement verification | App user ID, purchase history, product and entitlement status, locale/currency data, and Apple Ads attribution data when available. | Review |
| Superwall | Paywalls, subscription campaign logic, and paywall analytics | App user ID, device/app metadata, paywall events, subscription status, and onboarding attributes sent by Buff Kitty, including referral or promo code if you provide one. | Review |
| PostHog | Product analytics | Event names/properties, app lifecycle events, RevenueCat app user ID, and SDK metadata. Session replay is disabled in production builds in the reviewed app configuration. | Review |
| Supabase | Backend Edge Function and private rate-limit counters | AI request metadata, per-install device ID, RevenueCat app user ID, function logs, rate-limit counters, and transient AI analysis inputs. | Review |
| OpenAI | AI provider for protein estimation | Food descriptions or meal images when OpenAI is used for backend AI analysis. | Review |
| Google Gemini API / Google Cloud | AI provider for protein estimation | Food descriptions or meal images when Gemini is used for backend AI analysis. | Review |
| Open Food Facts | Public packaged-food barcode and nutrition lookup | Scanned product barcode, app version and developer contact in the request user agent, and routine network metadata such as IP address. | Review |
| Vercel | Hosting for these legal webpages | Routine request and hosting metadata for visitors to the legal site. | Review |