Sub-processors
Version 2026-09-26 · Effective September 26, 2026
Incorporated by reference into the Terms of Service. Terms · Privacy · DPA · Sub-processors
Joyful Innovation uses limited sub-processors to provide, secure, monitor, and support Atlas and related services. This page summarizes the categories of sub-processors that may process customer data when those services are enabled.
| Category | Purpose |
|---|---|
| Cloud hosting and infrastructure | Application hosting, databases, object storage, queues, backups, and observability. |
| AI model and embedding providers | AI generation, classification, embeddings, transcription, and related model operations. |
| Email, calendar, and communication providers | Transactional email, customer-requested inbox/calendar sync, notifications, and support workflows. |
| Payments and billing | Subscription billing, invoicing, tax support, checkout, and customer billing portals. |
| Analytics, diagnostics, and security operations | Service health, product analytics, error diagnosis, audit support, and abuse prevention. |
AI model and speech providers
These providers may process the prompts, files and audio you submit to Atlas's AI features. This list is generated from the provider configuration Atlas runs, so it names exactly the providers that can receive that data.
| Provider | Service | Purpose |
|---|---|---|
| Mistral AI | Mistral La Plateforme | Primary provider for AI generation, classification and related model operations. |
| Gemini API | Secondary provider for AI generation, classification and related model operations. | |
| Amazon Web Services | Amazon Bedrock, including the Anthropic Claude, Amazon Nova and Stability AI models it hosts | Fallback provider for AI generation, classification and related model operations. |
| Microsoft | Azure AI Foundry | Last-resort fallback for AI generation, classification and related model operations. |
| Deepgram | Deepgram speech API | Voice transcription and read-aloud audio. |
| Gemini API | Voice transcription and read-aloud audio. |
If your organization connects its own AI provider account, that provider processes your prompts under your own agreement with it and is not listed here.
For the other categories, customers may request the current named sub-processor list, planned changes, or security documentation by contacting privacy@joyfulinnovation.com.
This page is part of the Data Processing Agreement.