Data handling
How Pipevideo stores, processes, and retains your generation data.
What Pipevideo processes
When you submit a generation request, Pipevideo processes:
- Prompt text and chat message content
- Reference images and videos (URLs or data URIs you provide)
- Generation metadata — model, engine, provider, dimensions, timestamps
- Orchestration transcripts — LLM agent output used to compose the video project
Generated video storage
Completed videos are stored securely and served via signed URLs with expiration. The video_url in API responses and webhook payloads contains a time-limited download link.
API key security
API keys are hashed at rest. Pipevideo never stores or logs the full key value after creation. Only the key prefix (pv_ + first characters) is retained for identification.
Data retention
Generation records (status, metadata, costs, timestamps) are retained for usage reporting and billing reconciliation. Contact support@pipevideo.co for questions about data retention policies for your organization.
Third-party providers
Orchestration models may be served through the selected video orchestration provider. Prompt content is sent to that provider for the orchestration step. Review your compliance requirements when choosing orchestration models.
Webhook payloads
Webhook events contain generation ids, status, output URLs, and error messages. Configure webhook endpoints over HTTPS and verify Svix signatures. See Webhooks.
Your responsibilities
- Do not submit regulated personal data (PII, PHI) unless your compliance review permits it.
- Rotate API keys on a schedule and after suspected compromise.
- Store downloaded videos according to your organization's data policies.