Obsidian Security Platform

Obsidian Security is a SaaS and AI security vendor headquartered in Palo Alto, California. It addresses the risks created when employees, third-party apps and AI agents gain access to business-critical SaaS applications.

The platform covers AI agent security, shadow AI, MCP security and AI security posture management, together with SaaS posture management, threat detection, excessive privilege reduction and account takeover detection. Named customers include T-Mobile, Databricks, S&P Global and Snowflake.

• AI agent security: gives visibility, governance and runtime security for agents.
• Shadow AI and MCP security: detects unsanctioned AI and MCP connections.
• Posture management: monitors configuration of third-party apps.
• Excessive privileges: finds and reduces over-permissioned access.
• Threat detection: detects account takeover and access violations.
• Audit and compliance: reports on app security controls.

A Vietnamese bank can monitor Salesforce and Microsoft 365 for account takeover. A technology firm can govern which AI agents access SaaS data. An enterprise can reduce excessive privileges in SaaS admin roles.

Combines SaaS posture and threat detection with AI agent controls. A vendor-cited Forrester study reports 192% three-year ROI and payback under 6 months. Integrates with Microsoft 365, Google Workspace, Salesforce and Snowflake.

AI agents increasingly connect to SaaS apps with broad permissions; the vendor states over 70% of its customers let agents into third-party apps. Vietnamese enterprises adopting Salesforce, Workday and Microsoft 365 face similar governance needs.

Raised an $85M Series D on 2026-08-04 led by Crescent Cove Advisors at a $1.1B valuation. Headquartered in Palo Alto, California. Backers include Greylock, Menlo Ventures and IVP. Founding year is not stated on fetched pages.