Data retention policy
Sierra retains and processes Customer Personal Data for the duration of the applicable customer relationship and only as necessary to provide the Platform Services in accordance with the Agreement, DPA, applicable statements of work, documented customer instructions, and applicable law. Sierra maintains a data-retention schedule governing archival copies and their subsequent deletion.
Data archiving and removal policy
Upon expiration or termination of the Agreement, Sierra securely deletes Customer Personal Data through its standard deletion process. Before deletion, customers may request an export in a reasonably readable electronic format where technically feasible. Backup and archival copies are isolated, protected, and deleted in accordance with Sierra’s data-retention schedule, except where continued retention is required by applicable law. Sierra can provide written confirmation after deletion upon written request.
Data storage policy
Sierra hosts the Platform Services on AWS; Sierra’s corporate offices do not host Platform Services compute or storage. Customer Materials are logically isolated by unique tenant identifiers within Sierra’s multi-tenant architecture and are encrypted in transit and at rest, including backups, using industry-standard encryption. Encryption keys are managed and rotated using a hardware-security-module-backed cloud service. Access is restricted by role and least privilege, with multi-factor authentication for production and other privileged systems. Sierra performs periodic backups and replicates production data to a secondary region. Current subprocessors and their applicable countries or regions are documented in the Sierra Trust Center.
App/service has sub-processors
yes
Guidelines for sub-processors
App/service uses large language models (LLM)
yes
LLM model(s) used
OpenAI GPT, Anthropic Claude
LLM retention settings
Sierra has configured its approved LLM provider accounts for Zero Data Retention on the endpoints and models used by the Sierra Slack app. Prompts, contextual data, and model responses are processed transiently to generate the requested response and are n
LLM data tenancy policy
Sierra accesses LLMs through commercial API accounts operated by approved model providers.
Sierra isolates customer’s data using unique tenant identifiers and does not combine or share one customer’s prompts, context, or outputs with those of another
LLM data residency policy
Customer data submitted for LLM inference is processed by Sierra-approved providers in their applicable service regions. Current providers and processing locations are maintained in Sierra’s Trust Center: https://trust.sierra.ai.