Policy Hub
This page consolidates the AI subprocessors used by Phase Pal and explains how generative content is delivered to players.
AI Subprocessors
| Service Name | Purpose | Privacy Policy |
|---|---|---|
| Google Gemini | Text moderation and prompt-generation assistance for creator-facing AI workflows | https://policies.google.com/privacy |
| OpenAI | Image moderation for banner uploads and related safety checks | https://openai.com/policies/privacy-policy/ |
| Inworld AI | Primary provider for voice cloning and text-to-speech generation | https://inworld.ai/privacy |
| Replicate | Fallback AI provider for voice cloning workflows when legacy or configured flows use Replicate | https://replicate.com/privacy |
| Mem0.ai | Processing and storage for the memory feature | https://mem0.ai/privacy |
How Generative Content Is Delivered
In Phase Pal, generative content is delivered to players through seamless, real-time, voice-enabled conversations with our AI characters. This creates a deeply immersive and interactive experience.
- Player Interaction: The player speaks to an AI character using their microphone.
- Live Transcription: Our backend receives the player's voice, transcribes it into text in real-time, and sends it to our AI models.
- AI Response Generation: The AI generates a text-based response that is contextually relevant and in-character.
- Dual-Mode Delivery: The response is instantly made available in two ways:
- On-Screen Text: The AI's reply appears on-screen as dialogue (e.g., in a chat bubble or as subtitles).
- Synthesized Voice: The text is converted into high-quality speech using a TTS service, and the player hears the character speak the response aloud.