Data retention policy
Rippling retains Slack OAuth tokens, connection metadata, and operational logs only as long as necessary to provide and secure the MCP Gateway. Slack content is processed on demand and is not retained beyond operational requirements.
Data archiving and removal policy
When an integration is disconnected or deletion is requested, Rippling revokes OAuth tokens and removes associated connection data from active systems. Residual backups expire according to established retention schedules. Legally required security or audit records may be retained with restricted access.
Data storage policy
Credentials and connection data are tenant-isolated, encrypted in transit and at rest, and accessible only to authorized services and personnel. Rippling does not sell Slack data or use it to train general-purpose AI models.
Data center location(s)
United States
App/service has sub-processors
no
App/service uses large language models (LLM)
yes
LLM model(s) used
OpenAI GPT models and Anthropic Claude models, including Anthropic models accessed through Amazon Bedrock. Rippling AI may select between approved models based on the use case, performance, output quality, latency, cost, and safety.
LLM retention settings
Rippling’s third-party LLM providers are configured for zero data retention or an equivalent limited-retention configuration. Prompts, inputs, and model outputs sent to those providers are not durably retained after processing and are not used to train
LLM data tenancy policy
Customer data is isolated by company and access is enforced using the authenticated user’s company context and existing Rippling permissions. Only data authorized for and required by the requesting user’s interaction is provided for model processing.
LLM data residency policy
Rippling AI model processing currently occurs in the United States. Third-party model providers process data under Rippling’s Data Processing Addendum and applicable subprocessor terms.Rippling AI is not currently supported in Rippling’s EU data center