Third-Party Services
Staark uses third-party services to operate, secure, maintain, and improve the platform. This page explains the general purpose of each provider.
Effective Date: June 10, 2026
Vendors may process only the information needed to help Staark provide the service, depending on the feature being used.
Sign-in and account authentication
Types of information that may be processed
Name, email address, Google account identifiers, and authentication-related data may be processed when a parent signs in with Google.
Notes
Used when a parent or caregiver chooses Google sign-in.
Auth.js / NextAuth
Authentication session management
Types of information that may be processed
Session identifiers and authentication-related data may be processed to maintain account sessions.
Notes
Used to manage login sessions inside the Staark application.
Supabase
Database, application data storage, and profile image storage
Types of information that may be processed
Account information, child profile information entered by parents, intake responses, plans, reflections, messages, progress data, feedback, and uploaded profile images may be processed.
Notes
Used to store and retrieve user-owned platform data.
Stripe
Payments, subscriptions, billing, cancellation, and invoices
Types of information that may be processed
Billing contact information, subscription status, customer identifiers, invoice information, and payment metadata may be processed.
Notes
Staark does not store full payment card numbers.
Anthropic
AI plan and recommendation generation
Types of information that may be processed
Parent-provided context needed to generate recommendations, action plans, translations, or support responses may be processed.
Notes
Staark does not use parent or child personal information to train public AI models.
Vercel
Website hosting and deployment
Types of information that may be processed
Technical logs, IP addresses, device/browser information, request metadata, and deployment-related information may be processed.
Notes
Used to host and deliver the Staark web application.
HeyGen
Voice, avatar, and media-related features
Types of information that may be processed
Text submitted for speech generation, voice identifiers, generated audio or media response data, and media-related request metadata may be processed depending on the feature used.
Notes
Used for advisor voice and media experiences where enabled.
Deepgram
Speech-to-text and voice features
Types of information that may be processed
Audio submitted for transcription, transcript data, text submitted for fallback speech generation, and voice request metadata may be processed depending on the feature used.
Notes
Used for voice input and fallback text-to-speech features.
General Commitments
- Staark does not sell personal information.
- Staark does not use parent or child personal information to train public AI models.
- Staark uses vendors to operate, secure, maintain, and improve the service.
- Vendors should receive only information needed for their role in providing the service.