AI Providers & Data Processing
Effective starting: October 19, 2026
MaybeUndo uses selected third-party AI providers to support features such as content generation, transcription, voice, and editing assistance. This page explains the role of those providers and links to their current policies. Details about performance monitoring are provided below the table.
Our Terms of Service and Privacy Policy remain the governing documents for your use of MaybeUndo. Provider policies may change over time; the links below point to the provider's current information.
AI and monitoring providers
| Provider | Use in MaybeUndo | Provider policies |
|---|---|---|
| MaybeUndo | Coordinates AI-powered features and requests to external AI providers. | Privacy Policy |
| OpenAI | AI conversations, content planning, and assistance with video and presentation editing. | Data usage · Terms · Privacy |
| Google (Gemini API) | Generates videos from text prompts and any supporting context included in the request. | Terms and data usage · Privacy |
| ElevenLabs | Generates voiceovers and background sounds from submitted text, when requested. | Terms · Privacy |
| LangSmith | AI performance monitoring. Monitoring details | Terms · Privacy |
AI performance monitoring
LangSmith helps us monitor the reliability and performance of AI features. When enabled, this integration sends limited operational measurements: random request identifiers, timestamps, model names, response times, token counts, input-size measurements, and error codes.
It does not send prompt text, generated content, recordings, transcripts, email addresses, or account or project identifiers. Request identifiers help us connect performance measurements with our internal troubleshooting records.
Our LangSmith workspace is hosted in the United States and is configured to retain these traces for 14 days.
Questions
For questions about AI provider usage or data processing, contact us at support@maybeundo.com.