Aetherial Security

Background

Non-human identities dominate modern cloud environments, yet they operate outside traditional security controls and visibility model. This invisible attack surface, the Aetherial, is rapidly expanding, and houses leaked secrets, over-permissioned roles and orphaned accounts that can be silently exploited. Existing tools fail to connect activity across platforms leaving organizations blind to how these identities interact and create risk.

Overview

Aetherial Security is a centralized, cloud-native platform designed to deliver visibility and enforcement over non-human identities across GitHub and Google Cloud. It ingests identity data, access relationships, and activity logs through API-based, agentless collection, feeding a unified data pipeline that normalizes and enriches this information. Processed data is stored in two complementary systems: a knowledge graph (Neptune) that maps relationships between identities, permissions, and resources, and a search engine (OpenSearch) that enables rapid querying of logs and events.

On top of this foundation, the platform builds a dynamic security graph that exposes attack paths, such as a leaked GitHub token tied to a GCP service account with excessive permissions. Detection logic aligned to OWASP Non-Human Identity risks identifies issues like secret leakage, improper offboarding, and privilege escalation. The frontend dashboard provides a centralized dashboard where operators can visualize these relationships, investigate incidents, and understand risk in context.

Critically, the platform goes beyond detection by enabling enforcement. Through controlled API integrations, it can trigger actions such as revoking tokens, rotating secrets, or isolating compromised resources. This ensures that identity risks are not only identified but actively reduced, transforming fragmented cloud identity data into actionable security controls. The Aetherial is within reach.