Data retention policy
Contour AI, Inc. retains Customer Data, including messages and files from Slack channels a workspace admin has connected to a Contour project, for as long as that Slack connection and the associated Contour project or organization remain active. Contour does not apply a fixed retention window; data is kept to provide the service and is removed through the deletion paths below.
Data archiving and removal policy
Detaching a channel from a Contour project removes that channel’s derived knowledge and imported files from the project, and deleting a project permanently deletes the Slack content imported into it. Uninstalling the app or disconnecting the workspace revokes Contour’s bot token and deactivates every connected channel. Customers can request deletion of all of their organization’s data by emailing team@usecontour.ai.
Data storage policy
Customer Data is stored in Contour’s Amazon Web Services environment in the United States: Amazon RDS for PostgreSQL for messages, metadata, and derived knowledge, Amazon S3 for file attachments, and a per-organization vector index (Turbopuffer, AWS us-east-2) for search embeddings. All stores are encrypted at rest (AES-256) and in transit (TLS 1.2+), and Slack bot tokens are envelope-encrypted with AWS KMS.
Data center location(s)
United States
Data hosting details
Cloud hosted on Amazon Web Services: Kubernetes (Amazon EKS), Amazon RDS for PostgreSQL, and Amazon S3 in the US East (Ohio) region, with S3 replicated to US East (N. Virginia) for durability.
Data hosting company
Amazon Web Services (AWS)
App/service has sub-processors
yes
Guidelines for sub-processors
App/service uses large language models (LLM)
yes
LLM model(s) used
OpenAI models (currently GPT-5 family; versions may change), OpenAI text-embedding models, Mistral AI OCR, Anthropic Claude (may be used)
LLM retention settings
Contour’s LLM providers are configured to retain Customer Data no longer than their abuse-monitoring window (at most 30 days for OpenAI), or not at all under zero data retention. Data sent to a provider is used only to generate the response.
LLM data tenancy policy
Contour’s LLM providers operate in multi-tenant API environments, with each request processed in isolation under Contour’s own credentials. Within Contour, each organization’s data is isolated by organization ID and is not pooled across customers.
LLM data residency policy
Contour’s LLM providers process requests in their own infrastructure, primarily in the United States for OpenAI, and provider-side processing is not region-pinned. Contour’s own stored prompts, outputs, and embeddings, run in AWS US regions.