Cassandra AI

Crush Security Cassandra AI is a tenant-aware AI assistant designed for security operations teams. It allows users to query their organization's security environment using natural language, returning answers grounded in the organization's actual data rather than generic security knowledge. Key capabilities include: - Natural Language Queries: Users can ask questions in plain English about their security stack, coverage, subscriptions, and access controls. - Tenant-Aware Context: The AI operates within the context of the specific organization's environment, providing answers relevant to that tenant's configuration and data. - Cited Answers: Responses include citations and sources drawn from the organization's own data, rather than generalized or assumed information. - Permission-Based Access: The AI enforces role-based access controls, ensuring users only receive information they are authorized to view. - Contextual Conversation: The assistant retains conversation context and handles follow-up questions within a session. - Actionable Insights: Users can identify security coverage gaps, review subscription statuses, and assess overall security posture. Common use cases include: - Coverage analysis (e.g., identifying which services protect cloud workloads or where endpoint protection gaps exist) - Subscription management (e.g., tracking expiring contracts, total security spend by vendor) - Access and compliance queries (e.g., identifying admin access to a SIEM, reviewing access changes, checking SOC 2 compliance status) Security and privacy controls include full data residency within the tenant boundary, role-based response filtering, a full audit trail of AI interactions, and real-time data access rather than stale snapshots.