Security and Data Protection

Workgrid AI models are not trained on customer data. Conversations and user interactions are anonymized upon processing and reporting. All PII is obfuscated on Workgrid's side.

Customers have control over what systems they connect to Workgrid. For documents, customers have sole control over which documents the AI Assistant can index. For third-party systems, Workgrid uses API calls to retrieve specific data from the system and follows system permissions.

With Builder, Workgrid's no-code drag-and-drop editor, users have granular discretion over exactly which API endpoints to call from their source systems, ensuring control and understanding of exactly what data will be accessed, processed, and presented to end users.

Workgrid inherits user permissions from systems, ensuring that users can only access the data that they should, with minimal configuration.

Workgrid leverages industry leading pretrained models. In addition, customers can bring their own trusted LLM (BYOLLM), which enables customers to maintain AI governance frameworks and additional control over data and privacy.

Generative AI models used by Workgrid are configured with content filtering models that have been trained and tested to detect specific categories of potentially harmful content in both input and output.

By leveraging enterprise-grade services in our core conversational engine, we confidently inherit compliance from industry leading providers in addition to performing regular threat modeling to OWASP standards.

No, at this time Workgrid is not FedRAMP authorized.