An AI Assistant that understands SaaS security
Get clear contextual answers within seconds. Cut through data, surface what matters most for stronger posture and faster remediation, and act with speed.
Get clear contextual answers within seconds. Cut through data, surface what matters most for stronger posture and faster remediation, and act with speed.
Powered by a governed multi-agent system and context from the Obsidian Knowledge Graph, AI Assistant helps explain, prioritize and resolve critical SaaS security issues in seconds through natural language interaction.
The Obsidian Knowledge Graph connects users, apps, permissions, and activity data into a single, living map and Obsidian AI Assistant filters through all this data to get you what you need.
AI Assistant guides you through every investigation with a step by step reasoning process, pulling relevant data for posture, identity and activity. Every threat is examined, effectively helping you distinguish critical threats from false positives.
Whether it's posture violations or threats, Obsidian AI Assistant helps you take fast and governed actions. With guided remediation steps, built in guardrails and user approvals, your team can resolve issues faster while maintaining compliance and accountability.
Obsidian AI Assistant is built on the Obsidian platform and governed by Model Context Protocol (MCP), which orchestrates a suite of specialized agents. When you ask a question, Obsidian AI Assistant routes it to the appropriate agent(s) to pull real context from your environment to respond to your query. Every interaction is logged, auditable and grounded in real data.
Obsidian AI Assistant runs on a governed multi-agent system, each focused on a core area of SaaS security. Below are some categories of these agents:
Posture & Policy Management: Helping teams understand, enforce and refine security posture with clarity and control.
Posture Agent: Explains complex posture rules in simple language, detailing what they protect, their importance and maps them to compliance frameworks such s SOC2, ISO27001 and HIPAA.
Violation Agent: Detects, ranks and prioritizes posture violations based on business impact, data sensitivity and risk level.
Exceptions Agent: Manages exception requests with full justification, approval workflows, expirations, and audit trails to balance flexibility with governance
Threat Detection & Investigation: Streamline triage and accelerate investigations and reduce mean time to response
Threat Agent: Pull context from across the Obsidian platform with the Obsidian Knowledge Graph to guide analysts through triage and guided remediation. Overall, this reduces false positives and improve time to resolution
SaaS & AI Visibility and Control: Provide visibility into connected apps and tools, eliminating hidden risks and shadow usage
Shadow SaaS & AI Agent: Identifies sanctioned or unmanaged SaaS and AI applications, explains their associated risks and recommends remediation.
AI Guardian Agent: Flags over privileged or misconfigured agents and recommends right-sizing permissions to prevent risks or data exposure
Yes, Obsidian AI Assistant is governed, explainable and compliant by design. Every action runs within the MCP framework, ensuring all data access and agent actions are authorized and transparent. Obsidian enforces strict data handling aligned with security standards such as SOC2 and ISO27001 compliant, giving you full visibility and confidence in how AI operates within your environment.
Obsidian AI Assistant does not perform fully automated remediation. Instead, it provides guided, explainable recommendations to help you take the right actions quickly. All actions initiated through the Obsidian AI Assistant are explicitly approved and confirmed by the user, ensuring every action is governed and in your control.